international cartels developing · 4 the classic presentation of firms’ incentive to cheat on...
TRANSCRIPT
NBER WORKING PAPER SERIES
INTERNATIONAL PRICE-FIXING CARTELS AND DEVELOPING COUNTRIES:
A DISCUSSION OF EFFECTS AND POLICY REMEDIES
Margaret LevensteinValerie SuslowLynda Oswald
Working Paper 9511http://www.nber.org/papers/w9511
NATIONAL BUREAU OF ECONOMIC RESEARCH1050 Massachusetts Avenue
Cambridge, MA 02138February 2003
Margaret Levenstein is Associate Professor of Economics, University of Massachusetts, Adjunct AssociateProfessor of Business Economics and Public Policy, University of Michigan, and Faculty Research Fellow,National Bureau of Economic Research. Valerie Suslow is Associate Professor of Business Economics and PublicPolicy, University of Michigan. Lynda Oswald is Professor of Business Law, University of Michigan. An earlierversion of this paper “Private International Cartels and their Effect on Developing Countries” was written as abackground paper for the World Bank’s World Development Report 2001. We are grateful for comments fromSimon J. Evenett and participants in seminars at Yale University and the University of Massachusetts. Our thanksto Julia DeBeers, Arjun Jayadev, Jessica Nowicki, Natalia Lukito, Kurt Rademacher, and Sukmi Sukmiwaty forexcellent research assistance. Special thanks also go to Charlie Perkins of Pipe Logix for providing data and toDennis McNamara for assistance with chemical industry classifications. The views expressed herein are those ofthe author and not necessarily those of the National Bureau of Economic Research.
©2003 by Margaret Levenstein, Valerie Suslow, and Lynda Oswald. All rights reserved. Short sections oftext not to exceed two paragraphs, may be quoted without explicit permission provided that full creditincluding ©notice, is given to the source.
International Price-Fixing Cartels and Developing Countries:A Discussion of Effects and Policy RemediesMargaret Levenstein, Valerie Suslow, and Lynda OswaldNBER Working Paper No. 9511February 2003JEL No. L4, F1, O1
ABSTRACT
We examine the possible effects of private international cartels on developing countries bylooking in detail at three recent cartel cases, as well as at a broader cross-section of 42 recentlyprosecuted international cartels. We discuss the indirect effects on developing country producers,either as competitors or co-conspirators, as well the direct effects of cartels on developing countryconsumers. By combining trade data with a sample of US and European prosecutions ofinternational cartels in the 1990s, we are able to estimate the order of magnitude of the consequencesof these cartels on developing countries as consumers. In 1997, the latest year for which we havetrade data, developing countries imported $54.7 billion of goods from a sub-sample of 19 industriesthat contained a price-fixing conspiracy during the 1990s. These imports represented 5.2% of totalimports and 1.2% of GDP in developing countries.
Margaret Levenstein Valerie SuslowUniversity of Michigan Business School University of Michigan Business SchoolAnn Arbor, MI 48109-1234 Ann Arbor, MI 48109-1234and NBER [email protected]@umich.edu
Lynda OswaldUniversity of Michigan Business SchoolAnn Arbor, MI [email protected]
2
I. Introduction
The U.S. Department of Justice, the European Commission, and the Organization for Economic
Cooperation and Development have all recently voiced concern about, and in the former two
cases, increased their prosecution of, international cartels. These recent prosecutions of
international cartels in a wide range of industries demonstrate that cartels have pernicious effects
on consumers despite the obstacles created by legal prohibitions on collusion and individual
firm’s incentives to compete rather than collude.
As with private international cartels through history, most of the cartels recently caught in the
antitrust net of the U.S. or EU competition authorities are made up of producers in industrialized,
OECD countries. Therefore, it is not surprising that most prior studies of the impact of these
cartels focused on the better-documented effects on wealthy, industrialized countries.1 This
appears to be true of both business and public policy players, as there has been little activity on
the part of developing country governments or developing country consumers to respond to these
cartels even after they have been shown to exist. This contrasts with the actions of the Canadian
government, which has consistently pursued anti-competition cases against firms who have been
investigated first by either the U.S. Department of Justice or the European Commission. One
exception to this generalization is Mexico, which took action against the lysine cartel, and is
investigating the vitamins cartel.2 The lack of action in response to these cartels also appears to
hold true of private parties in developing countries, which have, with only a few exceptions,
apparently not actively sought civil remedies against cartel participants to the extent that
consumers in western, industrialized countries have. There are a variety of reasons – legal,
political, and economic -- why this may be the case. But, as this paper demonstrates, a lack of
impact on developing countries is probably not one.
1 ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT, see HARD CORE CARTELS, 6 (Paris, France 2000). See also speech by Mr. Mario Monti, Member of the European Commission in charge of Competition, Fighting Cartels: Why and How? Why should we be concerned with cartels and collusive behavior?, Address at the Third Nordic Competition Policy Conference (Stockholm, September 11, 2000) and the U.S. DEPARTMENT OF JUSTICE INTERNATIONAL COMPETITION POLICY ADVISORY COMMITTEE, FINAL REPORT (U.S. GPO Washington, DC 2000). 2 Brazil is also contemplating action against the lysine cartel. See Scott Kilman, European Commission Sets ADM Fine, WALL ST. J., June 8, 2000, at A4.
3
This, in turn, suggests that a more comprehensive approach to promoting competition may be
necessary. Current regulatory institutions are neither international enough nor sufficiently
focused on promoting competition rather than simply prohibiting particular anti-competitive
techniques to assure that global markets will be competitive and open to new producers. There is
currently no competition authority that considers it within their purview to assure that developing
country producers have access to markets uninhibited by restraints from private agreements by
established producers.
The extant research on the impact of cartels on developing countries focuses on commodity price
stabilization schemes among developing country producers of primary products. In these
studies, the analysis focuses on developing countries as producers and industrialized countries as
consumers. In contrast, the cartels in our sample produce sophisticated manufactured goods or
services; their members are largely international corporations based in industrialized countries.
We examine two aspects of the impact of these cartels on developing countries. First, we look at
developing countries as producers, either competitors to or collaborators with, these international
cartels. In three case studies, we examine the creation of barriers to entry by cartels and their
impact on developing country producers or potential producers. We also examine the methods
that may be used to induce cooperation with the cartel by developing country producers.
Second, we take a cross-section sample of all international cartels prosecuted by the U.S. and EU
in the 1990s and ask how price-fixing conspiracies may have affected developing country
consumers. This two-pronged approach gives a more complete picture of the varied direct and
indirect effects of international cartels on developing countries.
Section II begins with a brief overview of basic cartel theory and outlines the possible effects of
international cartel activity on developing country producers and consumers. Section III
describes three recent cartels and their effects: the citric acid, graphite electrodes, and seamless
steel tubes cartels. Section IV presents a cross-section sample of forty-two private international
cartels from the 1990s and estimates the effect of these cartels on developing country trade.
Section V discusses the legal issues posed by foreign plaintiffs seeking antitrust remedies in U.S.
courts for the anticompetitive conduct of international cartels. Section VI presents conclusions
and policy implications.
4
II. Cartels and Their Effects
A. Cartel Basics
There are a wide variety of organizations that may reasonably be described as international
cartels. The focus of this paper is explicit price-fixing or market division agreements, known in
policy circles as “hard core” cartels, among private producers from multiple countries. These
agreements are illegal in the U.S. and the European Union. They have been the focus of
increased prosecution by U.S. and European Union competition policy authorities over the last
decade. (They are illegal in many other countries as well, although laws and enforcement vary.)
There are other types of cartels, such as purely domestic cartels, private export cartels, and state-
run cartels. Our analysis is limited to private hard-core international cartels.3
Producers form a cartel with the goal of limiting competition. By restricting output and
increasing price, ideally to the price a monopolist would set (if the cartel controls the entire
market), profits will be jointly maximized. Assume for the moment that the firms in an industry
have overcome the coordination problems necessary to establish a cartel. Upon its creation a
cartel immediately faces the problem of how to escape from the Prisoner’s Dilemma: by raising
price above marginal cost, the cartel creates an incentive for each producer to cheat.4 Each firm
has an incentive to shave its price, increase its output and market share, and thereby increase its
profits. But if each firm did so, collusion would immediately dissolve into competition.
Repeated interaction (over time or across markets) can, by providing the incentive of future
collusive profits, deter firms from cheating in the present and allow them to escape this
Prisoner’s Dilemma. This tradeoff can be expressed very generally as requiring that the
3 Cartels with significant state involvement, such as OPEC, can certainly have important economic effects. Their goals, however, are much more complex than private cartels, including not only the maximization of joint profits, but economic stability and international political influence as well. The economic models that we use here, which presume a simple profit-maximizing objective function, are inadequate to address the functioning and impact of this set of international cartels. Thus, we have chosen to exclude them from our analysis. 4 The classic presentation of firms’ incentive to cheat on collusive agreements is George J. Stigler, A Theory of Oligopoly, 72 J. POLIT. ECON. 44 (1964). For further discussion of cartel economics and a survey of empirical research on cartel stability, see Margaret Levenstein and Valerie Suslow, What Determines Cartel Success?, University of Michigan Business School Working Paper, 02-001 (January 2002).
5
discounted expected stream of profits from future collusion exceed the profits earned by cheating
today:
E[Πcheating today] + E[Σ (Πfollowing an incident of cheating)δ] < E[Σ (Πcolluding forever)δ]
This means that the likelihood that an industry will bother to exert the effort to establish a cartel
will depend on several very basic factors that determine the expected profits associated with
colluding. These include the benefits of colluding, the benefits of cheating, the extent of
repeated interaction, and the discount rate.
Consider first the benefits of colluding. The incentive to create a cartel depends fundamentally
on the cartel members’ assessment of the potential for an increase in price to lead to an increase
in profits. This depends, in turn, on such factors as the price elasticity of demand (as demand is
more elastic, the potential for increasing profits decreases and the incentive to create a cartel
decreases); the rate at which future profits are discounted (as cartel members become more
impatient, collusion is harder to sustain); and variance in demand (at a minimum, demand
fluctuations create coordination and complexity problems, as the optimal price changes
whenever demand shifts).
Next, consider the benefits of cheating. If the benefits to cheating are sufficiently low relative to
the gains from colluding, the industry will find that it lies within the bounds in which collusion is
possible. In order to successfully collude, it will also be necessary for cartel members to devise
punishment mechanisms that provide necessary further deterrence to cheating. That is, the
second term on the left side of the inequality above (profits earned following an incident of
cheating), must be very low or even negative. The extent of multi-market contact -- the number
of times and places that two firms interact – is one determinant of the number of opportunities to
punish cheating; as it increases, so does likelihood that collusion will succeed. Another
important factor is industry structure on the buyer’s side of the market. If consumption is
concentrated in just a few customers, it is more likely that a cartel member would succeed in
6
increasing its market share substantially with a small cut in price and therefore be more tempted
to cheat. 5
The ability of the cartel to punish cheaters also affects the benefits of cheating. Antitrust
enforcement can make it difficult for a cartel to punish its members, particularly via a price war,
as such punishment would make its existence more obvious to the antitrust enforcement
authorities. Antitrust enforcement also limits the use of fines as punishments; in permissive
antitrust environments, cartels have frequently used mechanisms in which a firm that has sold
over its quota or in some other way violated the agreement (e.g., sold below the agreed upon
price or outside its assigned territory) simply compensates the other parties. These mechanisms
leave a trail of evidence, however, that must be avoided if there is a possibility of prosecution.
Following Stigler (1964), as developed by Green and Porter (1984), economists have focused on
the importance of the observability of cheating to collusive stability. When cheating cannot be
observed, it is harder to give firms an incentive not to cheat. It is more likely that collusion will
be disrupted either by cheating or by events that are empirically indistinguishable from cheating.
For this reason, firms in the cartel may find it useful to invest in information collection in order
to support the collusive equilibrium.6 In addition, cartel members often find that there is no
substitute for frequent face-to-face meetings, in order to compare market information and discuss
alleged occurrences of cheating.
Finally, having established sufficient incentives for existing industry participants to collude, the
ultimate critical element to sustainable cartel profitability is the existence of barriers to entry.
5 On the other hand, our sample of recent prosecutions of international cartels suggests that successful collusion is possible in industries with large customers. The vitamins cartel, for example, lasted many years and sold to very large customers. 6 David Genesove and Wallace P. Mullin, The Sugar Industry Learns to Organize Information Exchange, in LEARNING BY DOING IN FIRMS, MARKETS AND COUNTRIES 103 (Naomi Lamoreaux et al. eds., 1999) and Margaret Levenstein, Do Price Wars Facilitate Collusion? A Study of the Bromine Cartel Before World War I, 33 EXPLOR. ECON. HIST. 1107 (1996) examine the information collection procedures of two cartels in the sugar and bromine industries, respectively. Industry associations often engage in the collection and dissemination of information, which may facilitate collusion. The government may encourage this information dissemination, as the federal government did during the open price policies under the National Industrial Recovery Act, or as state governments did in the 19th century salt industry. See Barbara Alexander, The Impact of the National Industrial Recovery Act on Cartel Formation and Maintenance Costs, 76 REV. ECON. STAT. 245 (1994) and Margaret Levenstein, Mass Production Conquers the Pool: Firm Organization and the Nature of Competition in the Nineteenth Century, 55 J. ECON. HIST. 575 (1995).
7
When firms do manage to coordinate their conduct on incentive-compatible collusive strategies,
they create an incentive for outsiders to enter the industry. Coping with and preventing entry can
undermine the best-laid collusive plans. In some industries it may be that firms resist the
temptation to collude because they know that it would only lead to entry (which might, given any
cost of exit, make the incumbent firms worse off).
The next section outlines the possible effects that an international cartel could have on either
developing country consumers or producers. The price effects are clear enough. Therefore, we
focus on the hidden entry barriers that may be created as a necessary part of the cartel.
B. Potential Effects of Cartel Activity on Developing Countries
For developing country consumers, or consumers in any country for that matter, the direct cost of
a cartel is plain: price will increase if the cartel is successful. There may be other costs as well,
such as decreased product choice (if the cartelized product is differentiated and geographic
markets are allocated among producers) or a slower rate of technological change.
For developing country producers, again as with producers worldwide, there are both potential
costs and benefits. Developing country producers may benefit from an industry price umbrella
set by a U.S. or EU cartel, allowing non-cartel producers to sell at that price, or slightly below,
without having to adhere to a cartel production quota. There are, however, potential negative
effects as well, and developing country producers may be particularly susceptible to these
effects. In order to ensure cartel survival, international cartels may engage in activity that blocks
or slows entry by developing country producers. For example, cartel members may use tariff
barriers and antidumping duties to prevent entry by developing country participants.
International cartels may also use government-authorized, non-tariff barriers to prevent entry
(e.g., quotas or regulation) or punish outsiders (e.g., using trade reporting and import
surveillance by government agencies to track where other firms are selling). If these cartel-
imposed costs are significant, there will be a cost to the pace of economic development and the
development process.
In addition to those barriers intentionally or inadvertently provided by national governments,
cartels can also use private barriers to prevent entry. Historically, cartels have used a variety of
8
different techniques to block entry. These include the threat of retaliatory or predatory price
wars, use of a common sales or distribution agency (i.e., vertical foreclosure), and patent
pooling. For the most part, the public record on recent price-fixing cartels does not discuss
whether the cartel engaged in activities to block entry because such evidence is not necessary for
a criminal conviction, at least in the United States where price fixing is per se illegal. In part,
because of the secrecy surrounding cartel operations, we must rely largely on anecdotal evidence
from which only tentative conclusions can be drawn. However, we have found descriptions of
activities by contemporary international cartels that may have been attempts to deter or block
entry by developing country producers.
One example is provided by the price-fixing conspiracy in the EU steel beam market between
1988 and 1994. Steel makers who were colluding to fix the price of steel beams “restrict[ed] the
flow of information . . . in order to freeze out any new competitors," according to Karl Van
Miert, the EU competition commissioner.7 It is not clear from the published record what type of
information steel producers were trying to restrict in the steel beam case, but we do know that in
many industries information about technology and more formally, patent pools, have been used
by cartels in the past to create barriers to entry.8
Or, consider the actions of graphite electrode producers from the U.S., EU, and Japan between
1992 and 1997 (discussed more fully below). The U.S. Department of Justice alleged that
graphite electrode producers engaged in activity to disadvantage outsiders to their cartel,
claiming that they “agreed to restrict non-conspirator companies’ access to certain graphite
electrode manufacturing technology.”9 Again, while this charge appears in every individual
indictment, indicating it was agreed upon by all cartel members, the details of the firms’ actions
are not given.
7 Charles Goldsmith and Martin DuBois, European Commission Fines Steel Makers $116.7 Million, WALL ST. J. EUROPE, February 17, 1994, at 3. 8 See, for example, Steven W. Usselman, Organizing a Market for Technological Innovation: Patent Pools and Patent Politics of American Railroads, 1860-1900, 19 BUS. AND ECON. HIST. 203 (1990), and Leonard S. Reich, Lighting the Path to Profit: GE’s Control of the Electric Light Industry, 1892-1941, 66 BUS. HIST. REV. 305 (Summer 1992). 9 U.S. Dep’t of Justice, Japanese Subsidiary Charged with International Conspiracy to Fix Prices for Graphite Electrodes in the U.S., Press Release, February 23, 1998.
9
In another case, U.S. producers of ferrosilicon formed a cartel in 1989 and proceeded to use
antidumping laws in the U.S. and Europe to bar entry to non-cartel members.10 However, the
claim that firms were systematically filing anti-dumping claims and then withdrawing them
following a, presumably threat-induced, agreement with importers has not stood up to rigorous
empirical tests.11
These kinds of activities may be particularly effective in limiting entry from developing country
producers who are just entering international markets. If effective, developing country producers
may be excluded. Even after cartels are broken up, the existence of these barriers may force
developing country producers into joint ventures that limit their distribution or restrict it to
certain markets. Such joint ventures could then function as a way for colluding firms to
accommodate developing country entry into a cartel under their own terms or to engage in an
implicit cooperative pricing arrangement. These arrangements give developing country
producers access to the world market, but may do so at some cost to the degree of competition
that would otherwise obtain in the industry. In several recent international cartel cases, joint
ventures have been established in the years following the forced break-up of the cartel. This may
reflect an attempt to consolidate and restructure the industry in a more direct way, in light of the
break-up of the cartel.
Of course, both developing country entrants and established producers could also have other,
welfare-enhancing motives for establishing such joint ventures, such as sharing technology, local
market expertise, or capital. It is important to note that these explanations for joint ventures are
not mutually exclusive; a joint venture might well accomplish both welfare-enhancing and
competition-reducing goals of the participating firms. Joint ventures (and mergers) in industries
known to have a history of international price-fixing should be scrutinized by regulatory
authorities and structured so as to support the welfare-enhancing gains from cooperation while
10 Richard J. Pierce, Jr., Antidumping Law as a Means of Facilitating Cartelization, 67 ANTITRUST L. J. 725 (2000). 11 Using data from 1990 to 1997, Taylor (2001) finds that most withdrawn anti-dumping cases either have no effect on market price and quantity, or are followed by a decrease in price and increase in quantity. See Christopher T. Taylor, The Economic Effects of Withdrawn Antidumping Investigations: Is There Evidence of Collusive Settlements, U.S. Federal Trade Commission, Working paper 240 (August 2001).
10
allowing consumers in both developing and industrialized countries the benefits of enhanced
competition.
Given this overview of basic cartel operations and how cartel activity might affect developing
countries, we now turn to three contemporary international cartel cases to find specific
illustrations of these activities. In each case, we provide a brief overview of the industry and
then discuss the price-fixing conspiracy and its possible effects.
III. Three Illustrative Cartel Cases
As we will discuss in Section IV, there have been approximately forty international cartels
indicted and prosecuted in the 1990s by the U.S. Department of Justice and the European
Union’s European Commission. Of this sample, we have selected three cases to show the
potential effects of international cartels on developing countries: citric acid, graphite electrodes,
and seamless steel tubes. In these cases the effects of the cartel were felt worldwide and the
cartel exported a significant percentage of the product to developing countries. Also, although
the quality of the data varies, there is at least some price data obtainable in each case. There are
other contemporary cartel cases where the cartel, although international in its membership,
covered only a limited geographic scope (e.g., Western Europe). These cartels may have had
significant effects for a period of time, but are not ideal for our focus on developing countries. In
addition, there are other recent cartels in the sample that probably did have worldwide effects,
but were either services or customized products, so adequate price and trade data are not
available (e.g., cable-stayed bridges). In contrast, the three cases selected satisfy our criteria of
having both broadly felt effects and publicly obtainable price data.
A. Citric Acid Cartel
1. Industry Background
Citric acid is used primarily as a flavor enhancer and preservative, falling into a general category
of chemicals called acidulants. Acidulants are naturally occurring acids that inhibit the growth
of bacteria and can offset product sweetness with their tart flavor. In general, the main uses for
citric acid are in soft drinks (its largest end use), processed food, detergents, and pharmaceuticals
11
and cosmetics. The acidulant class also includes lactic, fumaric, malic and tartaric acids.
Fumaric acid, for example, competes against citric acid as a preservative. It is generally cheaper,
but has certain chemical characteristics (e.g., a stronger acid taste than citric acid), that make it
an inferior substitute for many processed foods.12 Citric acid is the most widely used acidulant,
accounting for about two-thirds of the total acidulant market.
Buyers can be large or small, but the large customers account for the bulk of citric acid sales.
Given that the greater part of citric acid production goes to beverage companies, such as Coca
Cola and Pepsi, the buyers are very large indeed. Procter & Gamble is also one of the largest
U.S. consumers of citric acid. In fact, in the United States, approximately 70 percent of citric
acid and sodium citrate sales go to 10 to 15 end users.13
There are two primary production processes – shallow pan and deep tank fermentation. The
deep tank process is preferred in most industrialized countries due to lower labor requirements
and better quality control. This process does, however, require large amounts of energy as an
input.14 Connor (1998) estimates the marginal cost of production at $0.60 per pound during the
conspiracy period.15 The shallow pan process is more labor intensive and less capital intensive,
and therefore operates on a smaller scale.
Production is concentrated in the U.S., Europe, and China, although there are citric acid
producers scattered throughout the world. In the late 1990s Western Europe, the U.S., and China
together had an 88% market share of world capacity, estimated at approximately 1.2 billion
pounds in 1994. Table 1 provides a summary of the key firms in the industry, their capacity, and
market shares. The U.S. industry in 1990, just prior to the start of the conspiracy, had three
players: ADM, Cargill, and Bayer AG (a German firm whose U.S. marketing was handled by
Haarmann & Reimer, its U.S. subsidiary). Cargill entered the industry in 1990, as the first
12 Fumaric Acid, CHEMICAL MARKETING REPORTER, July 24, 2000, at 33. 13 Petition for the Imposition of Antidumping Duties: Citric Acid and Sodium Citrate from the People’s Republic of China, filed by Akin, Gump, Strauss, Hauer, & Feld, L.L.P. with the U.S. Int’l Trade Comm’n, filed December 15, 1999, at 17. (Public version of document obtained from ITC website: http://dockets.usitc.gov.) Hereinafter referred to as “ITC Petition.” 14 ITC Petition at 15. 15 John M. Connor, What Can We Learn From the ADM Global Price Conspiracies?, Dept. of Agricultural Economics, Purdue University, Staff Paper #98-14, August 1998, at 11. Hereinafter referred to as “Connor (1998).”
12
producer vertically integrated forward from corn refining into citric acid production. All
producers are now vertically integrated. In Europe in the early 1990s there were five producers
in the citric acid market; the three largest were Bayer, Hoffmann-La Roche (a division of
Switzerland's Roche Holding), and Jungbunzlauer International AG (Switzerland). These
European companies, as well as smaller Chinese importing companies, satisfied most of the U.S.
import demand during the mid-1990s.
Chinese producers have presented the most vigorous competition to U.S. and European
manufacturers. Up to one hundred small firms entered the industry in the mid-1990s with the
help of the Chinese government. Although the exact price differential has varied, in general
Chinese citric acid sells in the U.S. for 10 to 20 cents less per pound than domestic supplies and
European imports. Some consumers consider China’s product to be lower quality and will not
consider buying it, despite the price difference. For others, particularly industrial users, price is
the major decision variable. Chinese exports peaked around 1994 and then dropped off as the
Chinese government withdrew its subsidies and raw materials prices increased. Exports from
China rebounded after the cartel was broken apart, suggesting that the effect of cartel barriers to
entry on limiting Chinese production was greater than the incentive effect of a price umbrella.
The Chinese producers as a group currently hold about 15% of the U.S. market share.
2. Price-Fixing Conspiracy and Its Effects
According to U.S. Department of Justice documents, firms in this industry fixed prices from
approximately July 1991 to June 1995.16 Although the citric acid cartel did not control world
production, it did account for 75-85% of sales in North America and Western Europe.17 Citric
acid firms have been convicted for these activities in the United States, Canada, and the
European Union. The fines are detailed in Table 2.
16 U.S. Dep’t of Justice, Justice Department’s Ongoing Probe Into Food and Feed Additives Yields Second Largest Fine Ever, Press Release, January 29, 1997. The reported cartel dates vary somewhat, depending on the particular firm charged and the antitrust authority or private plaintiff bringing the suit. Connor (1998), for example, notes that DOJ indictments filed against the European participants in the conspiracy list July 1991 to December 1996 as the cartel dates. See Connor (1998), at 11. 17 Connor (1998), at 13.
13
There have been several follow-on suits by customers claiming damages. One civil suit filed by
bottlers and food processors was settled in 1996 for a total of $94 million (ADM, H&R, HLR,
and Jungbunzlauer were defendants). Cargill was named in this civil suit, but exonerated. In the
court opinion of September 1, 1999 the judge wrote: “It is true that between 1990 and 1997
ADM, H&R, and Cargill always changed list prices within a month of one another and generally
did so in the same month…Although there appears to have been little competition in citric acid
list prices, Cargill did price aggressively in actual contracts.” This difference between list and
transactions prices is important to keep in mind when we look later at the price trends in the
industry in the past decade. In particular, large customers generally pay less than list price.
The members of the citric acid cartel fixed prices and allocated sales in the worldwide market,
issued coordinated price announcements, and monitored one another’s prices and sales
volumes.18 In addition, the cartel members recognized the importance of policing and enforcing
the agreement. They shared monthly sales figures and took stock at the end of the year of each
company’s total sales. A company selling more than its quota was required the next year to
purchase citric acid from a cartel member that was under quota.19
The structure put in place by the citric acid cartel members was quite elaborate. The senior
executives responsible for determining the broad outline of the cartel agreement were nicknamed
“the masters.” At first, when the cartel began in 1991, only the masters held meetings. Later, in
1993, “the sherpas” (lower-level executives) began to hold meetings as well in order to handle
the day-to-day workings of the cartel and work out grievances between members.20
The U.S. price trend from 1990 through 1999 is shown Figure 1. Two price series from two
different sources are plotted: Chemical Marketing Reporter (CMR) and Purchasing Magazine
(PM).21 One can see from the graph that the CMR data is more representative of a list price,
while the PM data reflects, at least to some degree, true transactions prices. Prior to the
18 U.S. Dep’t of Justice, Justice Department’s Ongoing Probe Into Food and Feed Additives Yields Second Largest Fine Ever, Press Release, January 29, 1997. 19 Kurt Eichenwald, U.S. Wins A Round Against Cartel, NEW YORK TIMES, January 30, 1997, at 1. 20 European Commission document DN: IP/01/1743. (05/12/2001) 21 Most of these data (1987-97) are taken from Connor (1998) Appendix Table 1, which presents price data compiled from various issues of CMR and PM. We have updated the data series from the same two sources through 1999.
14
conspiracy, during the time when the industry was adjusting to Cargill’s entry and Pfizer’s exit,
there was a price war. Prices in early 1991 were driven down to the high-50 cent range.22 The
price war ended in early 1991. List prices rose steadily after that, stabilizing at 85 cents per
pound in the U.S. between 1993 and 1996 (this is reflected in the CMR line). According to
Connor (1998), actual transactions prices, as reflected by the PM line in Figure 1, stayed from 1
cent to 5 cents lower than list prices. For example, in 1991 CMR reported that “despite the 68-
cent list price, agreements are currently settled at about 63 cents.”23
Although the transaction price increase is slightly less dramatic, both price series in Figure 1
show a steady increase in price and then a decline after the conspiracy ended. EU Competition
Commission Mario Monti reported that citric acid prices rose by 50 percent during the
conspiracy.24 One has to be careful, of course, about drawing strong conclusions from such
statements or from the price charts included in this paper, since they do not control for other
factors affecting price. For example, there are seasonal effects in pricing due to increased
demand from the beverage market in late spring and early summer.
More generally, charges of increased cartel prices must be interpreted with care because some
portion of the increase may reflect other factors such as rising raw materials costs or increases in
demand. The price charts are purely descriptive, and do not purport to control for other relevant
factors that may have affected prices during the conspiracy period. In addition, we do not
estimate what the price would have been in the “but-for” world. That is, although it is clear that
there was a conspiracy and that firms have admitted their guilt, we have not attempted to
estimate the competitive price or the price that would have prevailed absent the cartel. Any
conclusions, therefore, about the effects of cartel activity must be drawn with great care.25
In order to provide an estimate of the order of magnitude of the effect of this cartel on
developing country consumers, we start with Connor’s (1998) estimate that buyers in the U.S.
22 Melissa Shon, Cargill, Jungbunzlauer Slate Citric Acid Additions, CHEMICAL MARKETING REPORTER, March 30, 1992, at 7. 23 David Axinn, Citric Acid Marks Rise as Market Settles Down, CHEMICAL MARKETING REPORTER, July 22, 1991, at 18. 24 Competition: Monti Calls for Higher Fines on Cartels, EUROPEAN REPORT (September 13, 2000). Unfortunately, Monti does not specify whether he is referring to US or European price increases (or whether they were the same. 25 We are not able to estimate but-for prices due to lack of publicly available cost and transaction price data.
15
paid an extra 21-24% during the conspiracy, using marginal cost as the “but-for” or
counterfactual price.26 If we simply round down and assume that prices increased 20% on an
approximately 300,000 million pound per year market in the “Rest of the World” (i.e., markets
outside the United States, Canada, and Europe), so that the prices charged were ten cents per
pound above the competitive level (which is substantially less than the observed price increase),
this would amount to a cost of $30 million per year to consumers. The cost to these consumers
is also the benefit to cartel members. The monopoly profits earned in these markets provide an
incentive for firms to collude; no corresponding punishment or fine has lessened this incentive.
It is possible that developing country producers may have received an increased price during the
conspiracy period by riding on the coattails of the major producers. Conversely, developing
country producers may have been damaged if the cartel was able to somehow prevent imports
into its territory. The existing literature has reached no consensus on the net impact of these
effects on cartel outsiders. We do have some evidence of attempts to limit entry from citric acid
anti-dumping cases that were filed during the conspiracy period. India imposed anti-dumping
duties on citric acid imports from China in November 1998. Before the duties were imposed,
China had captured close to 40% of the Indian market for citric acid. If China was being
excluded from the U.S. and European markets, either through anti-dumping cases or private
restraints, they may have turned to India as an outlet for their product. Thus, this may not relate
directly to the cartel, but could be an indirect consequence. It is even possible that the multi-
national firms that participated in the cartel were able to influence Indian policy toward Chinese
imports.27
26 Connor (1998), at 10. Lawrence J. White, Lysine and Price Fixing: How Long? How Severe?, 18 REV. IND. ORGAN. 23 (2001). Lawrence J. White disputes Connor’s use of marginal cost as the “but-for” price for the lysine conspiracy. White argues that the true “but-for” price was higher, based on the fact that the market was a four-firm oligopoly that probably would not have converged at an equilibrium price at marginal cost. White also argues for a shorter cartel period than Connor. Of course, from a policy perspective, the relevant question is what is necessary to achieve a competitive price that assures an efficient allocation of producers' resources and individual consumption decisions. Thus, for our purposes, the marginal cost price is the relevant comparison. 27 There are other examples of attempts by international cartels to use anti-dumping laws to sustain collusion. The ferrosilicon price-fixing conspiracy lasted from 1989-1991 and involved producers from the U.S. and Norway. Five of the six major US manufacturers pleaded guilty and were fined. These same firms asked for, and received, anti-dumping duties that were placed on Brazil, China, and other countries. When the International Trade Commission found out about the U.S. firms' involvement in a cartel, it reversed the tariffs.
16
U.S. producers have twice tried to use the government to help protect the domestic industry from
Chinese imports. First, in 1995, while the cartel was still intact, producers lobbied the Office of
the U.S. Trade Representative to include citric acid on the list of various Chinese imports to be
hit with a high tariff. A last-minute agreement prevented the sanctions from being imposed.28
The second anti-dumping allegation was brought at the end of 1999 by ADM, Cargill, and Tate
& Lyle, who reacted to the rise in imports of citric acid from China by filing a petition with the
Department of Commerce and the International Trade Commission seeking anti-dumping duties
of 350% on Chinese imports. While U.S. prices in early 2000 averaged around 63-66 cents per
pound, citric acid from China was selling for about 53 cents per pound.29 According to claims
made in the case, the filing was prompted in part because two of the largest consumers of citric
acid, Proctor & Gamble and Ashland Chemical Inc. (a distributor) switched to Chinese citric acid
for their raw material needs. Contradictory testimony was given regarding whether the quality of
citric acid from China met U.S. standards. One Chinese supplier tried to qualify to supply
Quaker Oats, for example, and was turned down (although this same supplier does sell to smaller
U.S. food manufacturers). The ITC dismissed the case in February of 2000, after deciding that
there was no material injury.30 At the hearings, it certainly weighed against the producers that
these same producers had just been convicted and fined for cartel behavior. U.S. and European
governments must be extremely wary of such attempts by firms to use the state as a tool for
creating barriers to entry.
There has been rapid consolidation in the industry since the price-fixing conspiracy was
revealed. Internationally, Hoffmann-La Roche completed its sixth joint-venture facility in China
in 1997. Its partner, Wuxi Zhongya, is one of China’s three largest producers.31 Cargill and Tate
& Lyle are both investing in Brazil, where a high quality and low cost sugar supply is attracting
citric acid manufacturers.32 There are two ways to interpret these events. One is that, following
the demise of the cartel, western producers are accepting entry from developing country
28 Cheryl Cullinan Lewis, Citric Acid, PURCHASING, May 4, 1995. 29 Feliza Mirasol, DOC Investigates Possible Dumping of Citric Acid, CHEMICAL MARKETING REPORTER, January 17, 2000, at 4. 30 Clay Boswell, Pucker Up: A Taste for Tartness Drives Acidulants, CHEMICAL MARKETING REPORTER, May 29, 2000, at 16. 31 Kevin Gopal, Keeping the Faith, PHARMACEUTICAL EXECUTIVE, January 1, 1998, at 36.
17
producers. The alternative is that former cartel members are attempting to re-establish market
dominance, absent the cartel, through these joint ventures.
Prices have fallen, both in the U.S. and in Europe since the demise of the cartel. Figure 1 shows
the general downward trend in U.S. prices since late 1995. CMR and PM both report that prices
are down and stable, despite the fact that demand is strong. European prices, which tend to be
lower than U.S. prices, have followed a similar pattern: the average price per kilogram ranged
from $1.68 - $1.82 in 1995, $1.04 - $1.39 in 1997, and $1.06 - $1.17 in 1999 (still well above
Connor’s estimate of marginal cost).33
B. Graphite Electrodes Cartel
1. Industry Background
Graphite electrodes (GE) are large carbon columns used by electric arc furnaces (EAF) or “mini-
mills” in the making of steel. These mini-mills use graphite electrodes to generate the enormous
heat necessary to melt scrap metal and convert it back into a marketable steel product. GEs are
made from synthetic graphite, for which the primary raw materials are petroleum coke, coal tar
and petroleum pitch. The petroleum coke is crushed and mixed with the pitch into a paste, which
is then extruded through a press. The electrodes are baked and undergo a series of refinements.
The electrodes are then machined to meet the customer’s specifications.
GEs are the only material that can generate sufficient heat to melt scrap steel. There is no
competitive substitute, other than the more traditional methods of making steel (i.e., open hearth
and basic oxygen). GEs make up about 6-7 percent of the cost of converting scrap to steel.
Almost fifty percent of GE costs are raw materials costs, the bulk of which is petroleum coke
(also called needle coke for electrodes applications). Labor costs represent about twenty percent
32 Kiernan Gartlan, Tate & Lyle To Expand Brazilian Citric Acid Operations, DOW JONES COMMODITIES SERVICES, October 19, 2000. 33 Citric Acid, EUROPEAN CHEMICAL NEWS, March 6-12, 2000.
18
of total costs.34 The production process is highly electricity intensive, and therefore the
electricity portion of the cost varies by location within a country and across countries.
A new plant takes 3-4 years to build. A civil complaint filed by numerous steel producers
against the GE manufacturers highlights the significant barriers to entry that exist even without a
cartel: “The production of GEs is a mature, capital-intensive business that requires detailed
product and process know-how. It takes approximately four years to build a new plant with a
20,000-ton capacity. No significant new player has entered the industry since 1950.”35 Thus
entry is limited not only by high capital requirements but also by the importance of implicit
technical and market knowledge.
There was a shakeout and consolidation in the industry in the late 1980s and early 1990s, just
prior to the price-fixing conspiracy. The consolidation was precipitated by slumping steel
production. In fact, GE industry capacity has shrunk by one-third since the mid-1980s.36 The
number of producers has since stabilized. Table 3 provides a summary of the major firms in the
industry and their market shares. In this highly concentrated market, UCAR International of the
United States and SGL Carbon Corporation of Germany dominate, with a combined world
market share of roughly two-thirds. Both firms manufacture electrodes in many countries
(including such developing and transition economies as Brazil, Mexico, South Africa, Russia,
and Poland) and sell throughout the world. There are also a number of other firms who are not
global producers, but who do sell their product globally. The C/G Group, for example, has
plants only in the United States, but sells throughout the world. Supporting this world market are
fairly low transportation costs, generally less than 5% of the cost of the electrodes.37
34 Barbara Martinez, Robert Krauss Chairman CEO and President of UCAR International, DOW JONES INVESTOR NETWORK, October 6, 1995. 35 Ferromin International Trade Corp., et al. v. UCAR, et al. In the United States District Court for the Eastern District of Pennsylvania, Second amended complaint, filed May 1, 1999, at paragraph 47. Hereinafter referred to as the “Ferromin complaint.” 36 New Issues – UCAR International IPO, STANDARD & POORS EMERGING AND SPECIAL SITUATION NEWSLETTER, November 14, 1994, at 15. 37 Ferromin complaint, paragraph 50.
19
The share of EAF production as a percentage of total world steel production has grown rapidly
over the past two decades. Mini-mills now comprise about one-third of total steel production.
Table 4 shows how EAF production was distributed around the globe in 2000:38
2. Price-Fixing Conspiracy and Its Effects
Price-fixing by graphite electrode producers began in 1992 and continued through at least
1997.39 According to reports in the press, investigation of alleged price-fixing began after a
complaint from a steel manufacturer.40 Lawsuits and criminal charges have been brought in the
U.S., Canada, and the European Union.41 Convictions are detailed in Table 2. In the U.S., for
example, there were seven firms indicted for price-fixing, UCAR, SGL, C/G, Showa Denko,
Tokai, SEC, and Nippon, and six firms fined (C/G was granted leniency by the Department of
Justice). An eighth firm, VAW Aluminum, was fined by the European Commission, but not by
either the U.S. or Canada. Fines (not including civil damages) against these eight firms now
total almost $500 million.
After the GE firms pled guilty to the U.S. charges, dozens of civil suits followed. Almost forty
U.S. steel producers sued for damages, and many of these suits have been settled. In general,
although manufacturers in developing countries must have been damaged by many recent
worldwide cartels, such as the vitamin, citric acid, lysine, and steel tube conspiracies, they have
apparently not sued in U.S. courts. The graphite electrodes case is an exception, where a civil
lawsuit has been brought by a group of non-U.S. steel producers. 27 international EAF steel
producers, many of them from developing countries, filed the “Ferromin” antitrust suit in
February 1999.42 The plaintiffs’ firms reside in Turkey, Thailand, Australia, China, Australia,
and Sweden. The defendants named are UCAR, SGL, Tokai, C/G, Nippon and SEC. The
38 International Iron and Steel Institute, Trends and Indicators: World Steel Production Data, at http://www.worldsteel.org/trends_prod/prod06. (visited November 2000). 39 EUROPEAN UNION, Commission Fines Eight Companies in Graphite Electrode Cartel, European Union Press Release, July 18, 2001. As is the case in most such cases, the exact dates of the conspiracy are not known. The alleged dates of conspiracies vary depending on the claimant and the accused firm. 40 Adam Jones, Blowing the Whistle – American-Style, THE TIMES, February 24, 2000. 41 The Japanese Fair Trade Commission issued a warning to Japanese GE firms in March of 1998. There was no conviction or fine, apparently due to a lack of evidence. 42 One of the plaintiffs is the Ferromin International Trade Corporation, which is a U.S. company that purchased graphite electrodes on behalf of its Turkish affiliates.
20
plaintiffs claim that their purchases of GEs in the U.S., Europe, Australia and Asia totaled $180
million over 1992-97 and that they were overcharged an average of 45 percent during this period.
In June 2001 a U.S. District Court dismissed most of these claims giving standing only to those
plaintiffs who state they can show that the GEs they purchased were invoiced in the United
States.43 We will discuss the legal issues arising from this case and a similar Christie’s art
auction case in more detail in Section V.
The information that we have on the cartel structure and organization comes almost exclusively
from the U.S. Department of Justice’s investigation. Cartel members agreed to: 1) increase and
maintain prices, 2) eliminate price discounts,44 3) allocate volume among conspirators, 4) divide
the world market among themselves and designate the price leader in each region, 5) reduce or
eliminate exports to members’ home markets, 6) restrict capacity, 7) restrict non-conspirator
companies’ access to certain graphite electrode manufacturing technology, 8) exchange sales and
customer information in order to monitor and enforce the cartel agreement, and 9) issue price
announcements and price quotations in accordance with the agreement.
Each of the provisions listed above would be considered “normal” (necessary, but not sufficient)
for the successful operation of a cartel. One of the most interesting aspects of the conspiracy is
the agreement to restrict access to technology, although the Justice Department has, so far,
provided no details of these allegations. One of the most noteworthy absences, though, is a
provision of penalties for cheating on the agreement. This may have been implicit and discussed
in the meetings, but never formalized. Since they did collect and share information on sales for
the purposes of enforcing the agreement, there presumably would have been a discussion (or
implicit threats) of the consequences of cheating.
The alleged price increases by the cartel were significant. In the United States, graphite
electrode prices increased over 50% from May 1992 through February 1997. The Ferromin
antitrust claimants allege that the price increases they suffered averaged over 45%. In Canada
43 Ferromin complaint. 44 More specific information on this point is given in Government's Sentencing Memorandum and Government's Motion for a Guidelines Downward Departure (U.S.S.G. §5K1.1), U.S. Department of Justice, Filed October 19, 1999. It says that all forms of discounts were to be eliminated, including rebates and consumption guarantees.
21
prices rose by more than 90% over 1992-97.45 The Canadian market was much more
concentrated at the time, consisting only of UCAR and SGL, with a combined market share
during the conspiracy years of over 90 percent.
The U.S. price trend from 1980 through 2000 is shown Figure 2. The chart captures the fall in
prices during the steel slump of the late 1980s, a clear increasing trend in the nominal price of
GEs during the cartel period, and a decline after the firms were convicted by the DOJ. (The
dotted line indicates missing data for the mid-1980s.) Purchasing Magazine reports that the last
price trough was $2,100 per metric ton in early 1992.46 In May 1992 the U.S. price was $3,123,
and by February 1997 it had risen to $3,439.47
Developing country producers may have been able to increase their prices under the rising cartel
price umbrella. That does not mean that developing country producers would have charged the
same cartel price; there may be quality differences or other differences in transportation costs,
supply assurance, contract terms, and so on. Although this is a reasonable conjecture, given
profit-maximizing behavior on the part of developing country producers, we have no data to
corroborate this hypothesis.
Alternatively, developing country GE producers may have been damaged if the cartel was able to
prevent imports into its territory. Indian graphite electrode producers have made exactly this
accusation: “Producers claim that the electrodes are being dumped into India at a price of $2200
per tonne as against the international price of $3200 per tonne."48 In response to a complaint
filed by the Indian Graphite Electrode Manufacturers Association, the government imposed anti-
45 Industry Canada, Competition Bureau, Foreign Corporation Fined $12.5 Million for Price Fixing, News Release (July 18, 2000). 46 Hotline, PURCHASING, October 19, 1995. Viewed on the web at http://www.manufacturing.net/magazine/pu…ng/archives/z1995/pur1019.p5/102hots.htm. 47 There are a few details worth noting about the prices used for Figure 2. From 1992-97 they reflect prices in the United States, as laid out in the DOJ Sentencing Memorandum of October 19, 1999. Outside of that time period, prices are taken from a variety of sources, including Forbes, Oil and Gas Journal, Dow Jones Commodity Service, UCAR earnings reports, and C/G SEC filings. It is unclear whether the price quotes given before 1992 and after 1997 reflect world prices or U.S. prices. For example, one news source says “high-performance graphite electrodes are approaching $2,100 per metric ton” without specifically stating whether this price applied to only the U.S. or more broadly. 48 CVD on Graphite Electrodes Imports Likely, FINANCIAL EXPRESS, April 20, 1997, at 2.
22
dumping duties in 1997 on imports from the U.S., several European countries, and China.49
Since the anti-dumping claims were filed in 1996, while the conspiracy was still operating, it is
possible that the conspiring GE producers were trying to force the exit of, or at least discipline,
Indian producers.
There has been a clear downward price trend since the conspiracy ended, as shown in Figure 2.
This certainly reflects in part the Asian financial crisis that hit the steel industry and therefore the
graphite electrode industry in late 1998. There is some evidence, albeit anecdotal, that points to
readjustment to a new equilibrium in the industry since the cartel ended. One recent article
mentions a “market share-driven price war” that has cut prices by five percent.50 In addition,
individual companies have restructured in the face of mounting fines. Joint ventures are also
being formed. In 1999, for example, UCAR entered into a production and marketing joint
venture with Jilin Carbon, the largest Chinese producer of graphite electrodes.51 Whether such a
joint venture facilitates or controls Chinese entry is not yet clear, but it does suggest that
monitoring of industries by competition authorities after the breakup of a price-fixing conspiracy
may be warranted.
C. Seamless Steel Tubes (Oil Country Tubular Goods) Cartel
1. Industry Background
Seamless steel tubes, pipes, and casings are used in the construction of wells in the oil and gas
industry. They are often referred to in the trade literature as Oil Country Tubular Goods
(OCTG). Steel line pipes are used in the transmission of oil and gas from wells. Stainless steel
tubes, made by established steel producers, represent new competition for the traditional OCTG
product, and are the only substitute.
Demand is extremely variable over time. It is closely correlated with the amount of drilling
currently being undertaken by oil and gas firms, which in turn depends on the price of oil and
gas. This means that there is excess capacity in the steel tubes industry during periods of low oil
49 Sharad Goel, HEG, Graphite Not Elated Over Dumping Duty on Electrodes, THE ECONOMIC TIMES, June 16, 1997. 50 Purchasing Hotline, PURCHASING, June 1, 2000, at 3.
23
prices. During periods of increasing oil prices, oil producers seem to accept price increases in
OCTG, but OCTG prices also seem to come quickly down when the price of oil does, as steel
producers try to make use of existing fixed capacity.
Natural gas creates greater demand for OCTG than does oil drilling, because natural gas wells
are deeper. Thus, consumer substitution of natural gas for oil increases demand for OCTG (and
the reverse). OCTG costs are not a large enough portion of the cost of production to lead to a
shift between natural gas and oil in response to fluctuations in the price of OCTG.
Entry in to the industry is, in principle, fairly easy, and there are a large number of firms. The
industry structure is, however, both more complicated and much more concentrated than a
simple count of the number of firms would suggest (see Table 5). In the U.S., there are a small
number of firms that produce a full line of steel tubes, casings, and line pipes, and sell that line to
the industry. These firms are often vertically integrated steel producers. There are a large
number of smaller firms that produce less than a full line of OCTG products. These firms are
usually not vertically integrated and instead purchase semi-finished steel inputs. They often also
purchase some OCTG or line pipe products from other manufacturers in order to offer a full line
to their customers. These firms often customize (with specialized coatings, etc.) products for
their customers. The U.S. firms sell primarily or exclusively to the North American market,
which includes the Gulf of Mexico. Otherwise, U.S. firms do not seem active in the export of
OCTG.
Other leading producers are located in Japan, Germany, France, Italy, Argentina, Mexico, Brazil,
and Sweden. These producers sell to both U.S. and worldwide markets. Three large alliances,
including all of the members of the former cartel, dominate world trade. The largest alliance is
that of an Italian-Argentine firm (Techint), which controls OCTG producers in Mexico,
Argentina, Italy, and Canada. These various relationships among steel tube producers, either as
suppliers and customers, or as owners or partners in a joint venture, provide many opportunities
for cooperation and may substantially lessen competition in the industry from what one would
expect if these various industry participants were all independent competitors.
51 John E. Sacco, UCAR Enters Joint Venture with Jilin Carbon, AMERICAN METAL MARKET, October 17, 2000.
24
Internationally, the creation of alliances among major producers has also meant the consolidation
of their sales forces. In fact, in the case of the alliance among Nippon, Kawasaki, and Sumitomo
Metal it appears that the alliance is essentially the creation of a joint sales agency to distribute
their goods worldwide.52 The opportunity to combine its sales force with the existing
international sales network of Techint (DST) was apparently central to NKK’s decision to spin
off its OCTG unit to NKKTubes, which is now jointly owned by NKK and DST.53 While the
use of a single, consolidated sales network may provide efficiencies in distribution or
convenience for customers, it also make communication and coordination of prices and market
shares much easier, and effectively prevents cheating by firms who have delegated sales to the
joint distributor.
Smaller, independent oil and gas producers may rely on brokers while larger firms have in recent
years been more likely to establish direct, long-term relationships with OCTG producers. For
example, Pemex, the Mexican state-owned oil producer, entered into a long-term arrangement
with Techint, whereby the Techint group provides just-in-time supplies of OCTG allowing
Pemex to reduce its inventories to near zero.
This kind of relationship, which has grown more prominent since the demise of the cartel in
1995, has changed the structure of distribution in the industry. In doing so, it has increased the
competitive advantages associated with vertical integration and horizontal size, because being
large and diversified is necessary to being able to guarantee supplies to customers in an industry
with such high variance in demand. It also has increased barriers to entry as customers are tied
to long-term relationships.
2. Price-Fixing Conspiracy and Its Effects
It is unlikely that the cartel agreement had a direct impact on the U.S. market where prices are
above world levels because of anti-dumping tariffs currently in effect. There has been no
antitrust case to date in the United States. In December 1999, the European Commission fined
52 Audrey McAvoy, Japanese Steel Companies Discussing Seamless Steel Pipe Tie-Up, DOW JONES INTERNATIONAL NEWS (August 18, 1999). 53 NKK Merges Seamless Pipe Operations in JV with Grupo Techint, DOW JONES INTERNATIONAL NEWS (November 2, 1999).
25
four European and four Japanese steel manufacturers over $100 million, charging them with
fixing bids on seamless steel tubes and line pipes between 1990 and 1995. The European
manufacturers included the inventor of steel tubes, Mannesmann; British Steel, now Corus,
which exited the industry in 1994; Dalmine, indirectly owned at the time by the Italian
government but privatized in 1996; Vallourec, a French steel producer who specializes in tubular
products. The Japanese conspirators were NKK, Kawasaki, Nippon, and Sumitomo Metal.
These eight independent firms created a cartel organization called the “Europe Japan Club.”
Under the auspices of the Europe Japan club they agreed “that the domestic markets of the
different producers … should be respected” so that producers refrained from selling in the home
countries of the other members of the Club.54 In shared markets, the Club met regularly and
designated which company was to win a particular job by bidding an agreed upon price, with the
others to submit higher bids.
The European Commission decision covered restrictions on sales and pricing agreements in
Europe. According to the EC, the cartel agreement also restricted competition in “certain third
markets.” The fines issued by the EC did not reflect these non-European markets because, the
Commission concluded, there was no evidence that they had a restrictive effect on the European
Union. Further details on this agreement have not been made public by the European
Commission pending appeals by some of the accused. Because the EC has not included these
other “third markets” in its decision, it is likely that details regarding this aspect of the agreement
will never be made public. This points to an important weakness in international competition
policy. The competition authorities in Europe may well have information regarding restrictions
on competition in developing countries (or other developed countries), but under current law and
agreements there is often not permission, let alone responsibility, to share that information with
the affected parties.
We have detailed data for OCTG prices in the United States over the period in question.55
However, because of the substantial tariffs in place during this period, these may not be a good
proxy for worldwide prices. Average prices of OCTG are presented in Figure 3. (Note that
54 EUROPEAN COMMISSION, Commission fines cartel of seamless steel tube producers for market sharing, European Commission Press Release (December 8, 1999). 55 We are extremely grateful to Charlie Perkins of Pipe Logix for providing us with this data.
26
OCTG is produced in both electric resistance welded or ERW form and a seamless form. The
cartel exerted control over both products, and Figure 3 shows both price series.) The OCTG
price falls during most of the period of the conspiracy. However, this was also a period of low
and declining oil and gas prices. Thus the observed prices, even in the U.S. where the cartel’s
effect was presumably only indirect, may have been higher than they would have been under
competitive conditions. Further analysis, controlling for the price of oil and gas, is necessary to
obtain a quantitative estimate of the effect of the cartel on prices.
Price trends in the industry continue to mirror oil and gas prices. U.S. prices fell during the early
1990s, reaching a trough in mid-1995. They then increased for three years until declines in oil
prices in 1998 led to a 41% drop in U.S. OCTG demand and declining prices of OCTG. By the
middle of 1999, OCTG prices were again increasing, as they continued to do for most of 2000.
The share of worldwide seamless tube exports coming from Germany, France, England, and
Japan stayed roughly the same during the period of the cartel (Figure 4), actually increasing
slightly toward the end of the period. As the cartel included the major producers from each of
these countries, this measure is a reasonable estimate of cartel exports. To the extent that there
were alliances between the cartel participants and producers in other countries, this measure
actually understates the market share of the cartel. The fact that their market share does not
decline suggests that entry (or expansion by non-participants) was not a viable source of
increased competition during this period.
No evidence was found indicating that steel producers blocked entry or potential entry into the
OCTG market from developing country steel producers. Several of the participants have
production facilities in developing countries, including Brazil, Mexico, Argentina, and the
transition economies of Eastern Europe. It appears that any sales of OCTG by producers from
Eastern Europe will be accomplished through cooperation with one of these international
alliances.
Since the demise of the cartel, the industry has undergone a fairly substantial reorganization, in
which all parties to the cartel have joined in one of three international alliances. The largest of
these, with a 25% market share of world consumption of OCTG is led by the Techint, an Italian–
27
Argentinean firm controlled by the Rocca family. Techint controls Dalmine, the Italian member
of the cartel, Tamsa, a Mexican tube producer, and Siderca, an Argentine steel producer. They
are known jointly as the DST group. The Rocca family has been in the steel tube business since
before World War II. Tamsa is currently under investigation by the Mexican Federal
Competition Commission for taking advantage of its position as the sole seamless tube producer
in Mexico. There is no indication in published reports that this investigation is linked to the
European Commission charges.56 NKK, also a member of the Europe-Japan club, has now
formed an alliance with DST, as has a Canadian producer.
As mentioned above, the other three Japanese producers who were members of the cartel
(Nippon, Kawasaki, and Sumitomo Metal) have formed an alliance in which they use a single
sales agency to represent all three. Mannesmann and Vallourec, the other two firms in the
Europe-Japan Club have formed a joint venture to which they have transferred all their OCTG
production. They are also engaged in steel tube joint ventures with Corus, another member of
the Club that has exited the OCTG market.
China’s exports of seamless steel tubes have increased significantly, but the current focus of its
steel tube producers is improving manufacturing technology and product quality rather than
expansion of capacity for export.57 This suggests that, at least in the short run, China will not
significantly increase competition for established producers.
Tariffs continue to play a significant role in this industry and may well limit the entry of
developing country firms not aligned with one of the three groups that dominate the industry.
The European Union imposed anti-dumping duties on Ukraine and Croatia in February 2000. It
has had anti-dumping tariffs in force against six other East European countries since 1997.
Tariffs have been maintained in U.S. since 1995 against Mexico and since 1994 against Japan.58
The recently enacted steel tariffs specifically exclude OCTG from increased tariffs.59 Imports of
56 Mexico: Investigation into Tamsa, METAL BULLETIN, August 21, 2000, at 14. 57 Chinese Steel Industry Will Mainly Develop Flat Products, ASIA PULSE (September 20, 2000). 58 Sumitomo Anticipates OCTG Rebound, AMERICAN METAL MARKET (April, 5, 2001). 59 President of the United States of America, Steel Products Proclamation: To Facilitate Positive Adjustment to Competition From Imports of Certain Steel Products, (March 5, 2002) at http://www.whitehouse.gov/news/releases/2002/03/20020305-7.html.
28
pipe into the United States have increased in part because tariffs on steel sheets have encouraged
foreign producers to export finished products into the U.S.
D. Summary
There are a number of lessons to draw from these three cases. First, consider the potential for
consumer welfare effects. International cartels can clearly have worldwide effects, but until
now, no one has examined this issue directly. In certain cases these cartels significantly raised
prices for several years. These price increases were on sophisticated intermediate goods, which
are passed on to both consumers and downstream producers. It is particularly costly to the
development process to raise prices and limit entry on this set of goods. Without prosecution by
government authorities, consumers lack the information, resources, and, in some cases, legal
structure to protect their own interests.
Second, the cases give us several insights into firm behavior and follow-on policy implications.
One of the lessons is that access to technology and markets is actively limited by the cartels (or
at least such attempts are made), using both governmental (tariff) barriers and private barriers.
Governments must therefore be attentive to how they might inadvertently help to support cartel
practices. In addition, significant industry restructuring often follows the break-up of cartels, yet
no antitrust authority seems to be watching to see if competition is being preserved during the
restructuring.
Also, we observe that some of these cartels came together during periods of increasing price
competition, often following entry. In general, we suspect that cartel formation may also follow
periods of market integration. Market integration alone, without vigorous anti-cartel
enforcement, may give rise to increasing cartel activity rather than competition. Increasing
liberalization of international trade may have inadvertently, by increasing competition in
formerly protected national markets, increased the incentives for firms to participate in cartels.
Such a response undermines the process of international integration, and decreases the benefits
of economic integration to consumers around the world. It may also undermine political support
29
for international liberalization if citizens believe that private barriers to trade will simply replace
government-created ones.60
Finally, there is a regrettable lesson for empirical research. The effects of private (as opposed to
state run) international cartels on developing countries are quite difficult to determine, even on a
case-study basis. There is anecdotal evidence about prices and barriers to entry, but few
definitive conclusions can be drawn. There are enormous difficulties in estimating the
quantitative impact of cartels on developing country incomes because of the secrecy under which
cartels operate, the lack of antitrust prosecutions in developing countries themselves (leading to a
lack of information on the activities of cartels in developing country markets), and the general
lack of data on individual transactions that might have been influenced by the existence of a
cartel. For this reason, we turn to the trade data as a way to quantify, however roughly, the
effects of contemporary international cartel activity on developing countries.
IV. Estimate of Developing Country Trade Affected By Recent International Cartels
It is impossible to gauge the true number of international cartels in existence in the 1990s.
However, we do know that the U.S. Department of Justice and the European Commission have
recently successfully prosecuted at least forty-two different international price-fixing
conspiracies that were in force at some point in the past decade. The surge in U.S. prosecutions
of international cartels stems primarily from the revision and expansion of the Antitrust
Division’s corporate amnesty program in 1993. The number of corporations coming forward and
seeking amnesty rose from roughly one corporation per year to one per month.61 On the heels of
60 We should note that some of the cartels in our cross-section sample, discussed in the next section, clearly pre-date recent moves toward international liberalization. 61 Speech by Anne K. Bingaman, Assistant Attorney General, Antitrust Division, Opening Markets and Protecting Competition for America’s Businesses and Consumers: Goals and Achievements of the Antitrust Division, U.S. Department of Justice, Fiscal Year 1993 through March 1996, March 27, 1996, at 8. See also Howard Adler Jr. and David J. Laing, The Explosion of International Criminal Antitrust Enforcement, BUS. CRIMES BULLETIN: COMPLIANCE AND LITIGATION (March 1997). Adler and Laing state, for example, that “In 1991, only 1 percent of corporate defendants were foreign and no foreign individuals were charged that year. From July 1996 to January 1997, 20 percent of all corporations and 27 percent of all individuals charged were foreigners.” (p. 1)
30
this increased enforcement by the U.S., both the European Union as well as some non-European
countries have strengthened their anti-cartel laws and stepped up enforcement.62
From these recent international price-fixing cases, we have created a sample of forty-two
international cartels on which the cross-section analysis in this paper is based (Table 6). We
believe that this is close to the universe of international cartels that have been successfully
prosecuted by the United States or the European Commission for fixing prices during the 1990s.
Table 6 summarizes the dates of cartel operation, the legal entity (i.e., the U.S. or the EC) that
prosecuted the case, the country of origin of the indicted firms, whether firms from developing
countries are known to be participants in the price-fixing arrangement, and, finally, which
country or countries are known to be affected (as consumers) by the cartel. In order to appear in
this table, a cartel must satisfy the following five conditions: 1) it must involve more than one
producer (otherwise, we consider it an extension of monopoly power case); 2) it must include
firms from more than one country; 3) it must have attempted to set prices or allocate markets; 4)
it must have existed during part or all of the 1990s (so, for example, there are cartels in our
sample that began in the 1980s and ended in the 1990s); and, 5) it must have been successfully
prosecuted by the U.S. or EU (or both). This sample, like its intellectual antecedents, may be
biased as a result of its dependency on prosecution as a sample selection criterion.63
Table 7 shows reported market concentration figures and cartel price increase information for
selected cartels. The typical international cartel has operated in a highly concentrated market (in
those cases where we can find the information). Estimates of the increase in price resulting from
these cartels vary widely by industry. At the low end, for example, we have a price increase of
ten percent for the thermal fax paper cartel, which was formed as the industry was declining and
lasted for less than a year. At the high end there is the price increase estimate of 100 percent for
the stainless steel cartel, and 50-60 percent in the U.S. and 90 percent in Canada for the graphite
electrodes cartel.
62 See, for example, Michael Reynolds, EU Briefings, 18 INT’L FINANCIAL L. REV. 48 (1999). The article announces the decision within the European Commission to create a new unit to fight cartel activity. 63 See Richard A. Posner, A Statistical Study of Antitrust Enforcement, 13 J. LAW ECON. 365 (1970); George A. Hay and Daniel Kelley, An Empirical Survey of Price Fixing Conspiracies, 17 J. LAW ECON. 13 (1974); Peter Asch and Joseph J. Seneca, Characteristics of Collusive Firms, 23 J. IND. ECON. 223 (1975).
31
In order to determine whether developing countries were consumers of one of the cartelized
products in the sample, we matched the products in Table 6 with import-export data for the
sample period. The trade data come from Robert Feenstra’s, World Trade Flows, 1980-1997,
With Production And Tariff Data (Center for International Data, Institute of Governmental
Affairs, University of California – Davis, 1999). The data include trade flows (imports and
exports) for all countries, classified according to the Standard International Trade Classification
(SITC), Revision 2.64 The data include only trade in goods. The list of developing countries is
taken from the World Bank’s World Development Report 2000/2001.65
Tables 8 through 13 summarize import data for thirty-two of the cartelized products in Table 6
for 1997, the most recent year for which trade data are available.66 The sample size falls from
the forty-two to thirty-two for two reasons. First, the data on trade flows exclude services, so
cartels that fixed prices on services were ruled out for further analysis. Second, goods were
dropped from the sample where the data appeared to be misclassified or aggregated to such a
level that no reasonable match to the cartel product could be made. Tables 8 through 10 contain
a combination of less aggregated and more aggregated data. Whenever possible, the narrower,
4-digit SITC product code, was used to track the trade data, but if the data were missing for that
category, we then used the broader, 3-digit code to categorize the cartel product. Tables 11
through 13 use less aggregated classifications (4-digit SITC codes only), and the number of
products for which we can obtain data falls to nineteen. Even then, the SITC codes are,
unfortunately, often broader than the product affected by cartel behavior. We indicate
64 Countries that were formerly part of the Soviet Union are conspicuous by their absence from World Trade Flows. Thus, the data on imports, exports, and Gross Domestic Product presented here simply exclude those developing countries that were formerly a part of the Soviet Union. There are also cases where World Trade Flows grouped smaller countries together (especially smaller island countries). We do not believe that this leads to any substantial misclassification in the data presented here. 65 According to the World Bank’s classification, there are 155 developing countries, divided into three groups (low-income, lower-middle-income, and upper-middle-income) and 52 “high income” countries. Examples of “low-income” countries are Armenia, India, and Vietnam; examples of “lower-middle” are Albania, China, and Thailand; and, examples of “upper-middle” are Argentina, Czech Republic, and Turkey. Gross Domestic Product (GDP) figures are calculated from World Bank data (www.worldbank.org/data/countrydata/countrydata.html). The World Bank provides detailed data on its website with country-specific statistics. The figures for total GDP by country categories are based on the same set of countries as those in the World Trade Flows sample. 66 These are the products for which we have been able to find minimally reliable data in international trade statistics. These data problems are discussed further below.
32
discrepancies between the cartelized products and the SITC categories in the notes to Tables 8 –
13.
Tables 8 and 11 report 1997 imports of “cartel-affected” products as a percent of total imports to
developing countries (using more and less aggregated product definitions, respectively) with
countries aggregated by income categories. Tables 9 and 12 present 1997 import data for these
same products, showing them as a percent of total GDP. Tables 10 and 13 give the total dollar
value of cartel-affected imports. We also report in each of these tables, just for comparison, the
analogous import values for high-income countries. In all of these tables we have deleted the
cartels that dealt with services, but we have left in the names of each of the thirty-five cartel
goods, so that it is easy to see where the missing data problems occur. For example, in Tables 8-
10, we do not even have 3-digit data for explosives, nucleotides, or zinc phosphates (thus, the
sample size falls to thirty-two). The number of products with missing data grows to sixteen in
Tables 11-13, where we try to find 4-digit SITC category matches.
Although Tables 8-10 cover a larger sample size, the product definitions are often much too
broad. We therefore focus on Tables 11-13. Examining the sub-sample of nineteen products –
those products that were cartelized at some point during the 1990s and for which we were able to
obtain a good match to the trade data – the total value of such “cartel-affected” imports to
developing countries was $54.7 billion (see the last row in Table 13). This figure made up 5.2%
of all imports to developing countries in 1997 and equaled 1.2% of their combined GDP. The
impact appears to be largest on the most developed countries of the developing world. Cartel-
affected imports made up 5.6% of imports and 1.3% of GDP for the “upper middle income”
countries who have the income and industries that demand and rely on imports of sophisticated
intermediate manufactured goods. While the total value of cartel-affected imports is higher for
high-income countries ($157.9 billion compared to $54.7 billion), these imports represent a
smaller proportion of imports and GDP (4.6% and 0.9% respectively). Of course, in countries
where producers belonged to the cartel, domestic production as well as imports is affected by
cartel behavior.
Using more aggregated data, we can obtain data on trade for almost all industries in which firms
have been convicted of fixing the price of goods (32 of 35 industries). This sample is more
33
comprehensive, in that it includes all cartelized goods, but it is also less accurate, because the
data include trade in many products that were presumably not affected by cartel behavior. Thus,
these figures present an upper bound to the value of affected trade in these industries. This
upper bound for the total value of affected trade is $155.9 billion of developing country imports,
representing 12.2% of their imports and 2.8% of their GDP (Tables 8 - 10).
These numbers clearly do not represent an exact value of the imports to developing countries
affected by all international cartels. The estimates are, on the one hand, biased downward
because we include only some of the forty-two known price-fixing conspiracies. At the same
time, even our lower estimate of affected trade, including only nineteen products, includes cases
where the trade categories are broader than the products whose prices were fixed by the cartel.
In general, when interpreting the trade data it must be kept in mind that some of the cartel
product-SITC matches are poor. These estimates are intended to give a sense of the order of
magnitude of affected trade, not an exact measure.
We can illustrate the typical problems with the data by using the citric acid, graphite electrodes,
and seamless steel tubes cartels again as illustrative cases. We have not been able to obtain
accurate international trade data for citric acid. Therefore, citric acid simply does not appear in
Tables 11-13. As with other narrowly specified chemicals, we suspect that there is
misclassification in the trade data, but whatever the source of the problem, volume of trade in
citric acid is simply not available. Even if we did have import measures they would understate
the full impact of the cartel on developing country consumers who pay higher prices not only for
raw citric acid, but also for a wide range of citric-acid containing goods. We can, however,
obtain data for the broader category of “carboxylic acids and their anhydrides and halides.” This
latter category is so broad that it contains at least five different products that have been affected
by cartels. We use data on trade in “carboxylic acids and their anhydrides and halides” in Tables
8 – 10, but there are surely other products in that category in which there is no cartel activity.
For graphite electrodes the situation is slightly better. Table 11 suggests that graphite electrodes
constitute a significant fraction of developing country imports (0.95%). Graphite electrodes are
important to developing countries that manufacture steel using the EAF process, but the data that
we present here undoubtedly vastly overstates their value, as this import category includes not
34
only graphite electrodes, but all “otherwise unclassified electrical equipment.” This broader
category again includes more than one cartel product, as carbon cathode blocks also fall into this
same catchall. On the other hand, if we did have a good match to graphite electrodes, that
measure would understate the impact of the cartel on trade because the product is an
intermediate good that is also imported into the developing country in a more processed state.
To the extent that the graphite electrode cartel increased the price of steel imports to developing
countries, focusing on graphite electrode imports understates the impact of the cartel. Therefore,
the data must be read with a degree of skepticism.
The OCTG cartel-trade data match is also problematic. This category (seamless tubes and pipes;
blanks for tubes and pipes) is much broader than the oil and gas goods that were included in this
particular conspiracy. However, there have been recent European Commission decisions
convicting an overlapping set of steel producers for fixing the price of steel heating pipes, steel
beams, pre-insulated pipes, and stainless steel during the late-1980s to mid-1990s.67 Thus, it is
possible that the prices of the other steel pipe products included in these import data have been
affected by these various activities. Some imports included in these figures were certainly
produced by firms who were not a party to these agreements. However, given the substantial
market shares of the firms in the cartel, it is likely that their behavior changed the prices charged
by firms who were not a party to and not even aware of the price fixing of their larger
competitors. Without more information about the secret activities of cartels, it is impossible to
determine the quantitative effect of these cartels on developing country incomes.
Even with these qualifications, it is clear from the magnitude of these figures that cartels have
adversely affected a not insignificant portion of the trade, and therefore the trade balance, and
consumption of developing countries. (Following the industrial organization literature, we focus
on trade and consumption, though the impact on the trade balance is not an unimportant issue in
a period in which some developing countries have experienced severe currency crises.) Given
the actual and potential effects on trade that reach into the tens of billions of dollars, a natural
67 Charles Goldsmith and Martin DuBois, European Commission Fines Steelmakers $116.7 million, WALL ST. J. EUROPE, February 17, 1994, at 3; Emma Tucker, European Commission Ten Companies Penalised for Fixing Prices of Insulated Steel Heating Pipes, FINANCIAL TIMES, October 22, 1998, at 3: and Philip Burgert, EC Issues Fines for Stainless Price Fixing, AMERICAN METAL MARKET, January 26, 1998, at 2.
35
question to ask is why these many affected countries are not seeking damages from cartel
member firms in their home countries. In particular, given that the United States has the
strongest laws and enforcement record against price-fixing, a legal mechanism for civil suits
(which, the European Union, for example, does not have for antitrust violations), and some of the
richest companies, why is it that there are relatively few lawsuits brought by foreign companies
seeking damages? Not surprisingly, given the many recent international cartel prosecutions,
there are a number of parties interested in this issue. The question of whether U.S. antitrust law
can be applied to foreign transactions has recently become a lively legal issue, and it is one that
we turn to now.
V. Foreign Plaintiffs’ Access to U.S. Courts in International Cartel Cases
Consumers in developing countries harmed by the activities of international cartels may be
unable to pursue legal remedies in their own countries, either because domestic antitrust laws
prohibiting such behavior do not exist, do not provide adequate remedies, or are not enforced by
the relevant authorities. In many such instances, those consumers may look to U.S. law instead
for remedies for the antitrust injuries they have suffered. The ability of such plaintiffs to sue in
U.S. courts is restricted, however, by the fact that U.S. antitrust laws do not reach all
anticompetitive conduct. Rather, U.S. antitrust laws apply to “anticompetitive conduct directed
at foreign markets that directly affects the competitiveness of domestic markets,” but not to
“anticompetitive conduct directed at foreign markets that only affects the competitiveness of
foreign markets.”68 In determining whether U.S. antitrust laws will apply to specific acts, the
courts look to the effect of the anticompetitive conduct, not the situs of that conduct.69 This
“effects” test can thus subject defendants whose anticompetitive conduct occurred solely outside
the U.S. to suit in U.S. courts, where the effects of that conduct are felt in the United States.
Historically, U.S. courts have not been particularly sympathetic to the claims of foreign antitrust
plaintiffs whose claims arise from anticompetitive conduct directed at foreign markets.
However, a March 2002 decision of the U.S. Court of Appeals for the Second Circuit, Kruman v.
68 Kruman v. Christie’s International PLC, 2002 U.S. App. Lexis 3895 at *17 (2d Cir. Mar. 13, 2002). 69 United States v. Aluminum Co. of America, 148 F.2d 416, 433-44 (2d Cir. 1945).
36
Christie’s International PLC,70 has suggested a broader mechanism by which foreign plaintiffs
can pursue legal remedies in the U.S. courts for the anticompetitive behavior of international
cartels.
The access of foreign antitrust plaintiffs to U.S. courts is largely governed by the Foreign Trade
Antitrust Improvements Act (FTAIA) of 1982,71 which Congress enacted in an effort to clarify
application of U.S. antitrust laws to foreign conduct and to limit application of U.S. antitrust
laws when non-import foreign trade is involved. In particular, Section 6a of the FTAIA
provides, in relevant part, that the Sherman Act “shall not apply to conduct involving trade or
commerce . . . with foreign nations unless . . . such conduct has a direct, substantial, and
reasonably foreseeable effect . . . on trade or commerce which is not trade or commerce with
foreign nations and such effect gives rise to a claim under” the Sherman Act.72
Until the Second Circuit’s recent decision in Kruman, all of the federal courts which addressed
this issue agreed that the FTAIA requires foreign plaintiffs suing under U.S. antitrust law to
show: (1) that the alleged anticompetitive behavior had a “direct, substantial and reasonably
foreseeable effect” on the U.S. marketplace and (2) that an anticompetitive effect on the U.S.
marketplace gave rise to the plaintiff’s claimed injuries.73 The coupling of these two
requirements effectively bars many foreign plaintiffs from suing in U.S. courts for international
cartel activities. To proceed, plaintiffs must be able to show that their injuries were caused
specifically by the anticompetitive effect of the defendant’s conduct on the U.S. marketplace and
not by anticompetitive conduct that affects a worldwide market, even if that market includes the
United States.
In Ferromin International Trade Corp. v. UCAR International, Inc.,74 for example, 27 plaintiffs
had alleged that they suffered injury as a result of price fixing and market allocation in the
worldwide market for graphite electrodes between 1992 and 1997. The U.S district court
70 2002 U.S. App. Lexis 3895 (2d Cir. Mar. 13, 2002). 71 Pub. L. No. 97-290, 96 Stat. 1246 (codified at 15 U.S.C. § 6a). 72 15 U.S.C. §6a. 73 See, e.g., Den Norske Stats Oljeselskap As v. HeereMac Vof., 241 F.3d 420 (5th Cir. 2001); Ferromin Int’l Trade Corp. v. UCAR Int’l, Inc., 153 F. Supp.2d 700 (E.D. Pa. 2001); In re Copper Antitrust Litigation, 117 F. Supp. 2d 875, 876 (W.D. Wis. 2000); de Atudcha v. Commodity Exch., Inc., 608 F. Supp. 510 (S.D.N.Y. 1985). 74 153 F. Supp.2d 700, discussed supra notes ___ and accompanying text.
37
dismissed the claims of 16 of the plaintiffs, stating that although the plaintiffs had alleged a
number of anticompetitive effects upon the U.S. marketplace resulting from the defendants’
conduct, the plaintiffs had not alleged that their injuries stemmed from the effect of the higher
prices for graphite electrodes in the U.S. market (as opposed to higher prices in other, foreign
markets). Indeed, the court found that the higher prices paid by those foreign plaintiffs were
caused by anticompetitive effects in foreign countries, not in the U.S. The court allowed the
claims of the remaining 11 plaintiffs to go forward, however, because some of the electrodes
purchased by these plaintiffs were invoiced in the U.S., thus satisfying the “causal requirement”
that these plaintiffs were injured as a result of higher prices in the U.S. marketplace.75
In its recent decision in Kruman, however, the Second Circuit deviated from prevailing
precedent on the meaning of Section 6a of the FTAIA, opening the door to more suits by foreign
plaintiffs. In a sense, the Kruman decision was a narrow one, as it was based specifically upon
existing Second Circuit precedent, and thus its impact on the law of other circuits is uncertain.
In another sense, however, the Kruman decision was of wide-ranging impact, both because of the
prominence of the Second Circuit and because the decision created a circuit split on an issue of
significant importance – the availability of U.S. antitrust remedies to foreign plaintiffs who were
harmed by anticompetitive conduct directed at least in part to a foreign marketplace. The legal
and policy implications of the Kruman decision are substantial and deserve careful exploration,
as the issue may well end up before the U.S. Supreme Court for ultimate resolution.76
The Kruman court rejected the two-step test articulated by all other federal courts addressing this
issue. Specifically, the court determined that the FTAIA had not altered prior law in the Second
Circuit with respect to the nature of the effect that anticompetitive conduct directed at foreign
markets must have on the domestic marketplace in order to be actionable under U.S. antitrust
laws. Under the Second Circuit’s National Bank of Canada rule,77 “anticompetitive conduct
directed at foreign markets is only regulated by the Sherman Act if it has the ‘effect’ of causing
injury to domestic commerce by (1) reducing the competitiveness of a domestic market; or (2)
75 Id. at 706. 76 A detailed analysis of the Kruman decision is outside the scope of this paper, but we plan on addressing this issue in future research. 77 See National Bank of Canada v. Interbank Card Assoc., 666 F.2d 6, 8 (2d Cir. 1981).
38
making possible anticompetitive conduct directed at domestic commerce.”78 The first prong of
this test encompasses anticompetitive conduct that is directed toward both foreign and domestic
markets and that actually reduces the competitiveness of the domestic market. The second prong
encompasses anticompetitive conduct that is “directed only at a foreign market, but has the effect
of allowing a separate course of conduct that directly affects the competitiveness of . . . domestic
markets.”79
The effect of the Second Circuit’s rule is to allow antitrust suits to proceed in instances in which
other federal courts would find the FTAIA prohibited such suits. The plaintiffs in Kruman were
buyers and sellers at foreign auctions who alleged that they were overcharged for auction
services in auctions held outside the U.S. as a result of a price-fixing conspiracy by the
defendants. (A separate suit was brought by plaintiffs who alleged they had been overcharged
for auction services in auctions held in the U.S. as a result of domestic price-fixing conspiracy by
the same defendants.) Because the plaintiffs had alleged only that they had paid inflated prices
at foreign auctions, the district court dismissed their claims, stating that they had not satisfied the
FTAIA’s requirements. While the imposition of high prices overseas may have had an effect
ultimately on the U.S. marketplace, the plaintiffs had not alleged that this domestic effect gave
rise to their injuries.80
The Second Circuit reversed, noting that the plaintiffs had alleged that the domestic price-fixing
scheme could not have succeeded in the absence of the foreign price-fixing scheme. If this was
true, the foreign conduct clearly had an anticompetitive effect upon the domestic market. The
“conduct” at issue could be described as an agreement to fix prices in both foreign and domestic
markets, which conduct clearly has an effect upon domestic markets because it includes conduct
directed at a domestic market. Alternatively, the “conduct” could be “described as an agreement
to fix prices in the foreign auction market that made possible an agreement to fix prices in the
domestic auction market.”81 In either event, one of the prongs of the National Bank of Canada
rule is met, and the FTAIA would not bar the plaintiffs from bringing suit in a U.S. court.
78Kruman, 2002 U.S. App. Lexis 3895, at *5-*6. 79 Id. at *22. 80 Kruman v. Christie’s Int’l PLC, 129 F. Supp.2d 620 (S.D. N.Y. 2001). 81 2002 U.S. App. LEXIS 3895 at *40.
39
Without ruling on the issue, the Second Circuit discussed the applicability of its ruling to
international cartels. In identifying the type of conduct that would satisfy the first prong of its
National Bank of Canada test (i.e., conduct that would have the effect of injuring domestic
commerce by reducing the competitiveness of a domestic market), the court provided the
example of an international cartel whose anticompetitive behavior was directed at both domestic
and foreign markets, and stated that “the FTAIA was clearly intended to regulate such
conduct.”82
The Kruman court did limit the scope of the conduct that could be regulated by U.S. antitrust
laws. The defendants had argued that given the nature of the modern global economy, with its
global marketplaces, any anticompetitive conduct that affects foreign markets could conceivably
affect the U.S. economy and so lead to a suit under U.S. antitrust laws. The Kruman court
rejected this line of reasoning, however, noting that under its test, such conduct would be
actionable only if it caused injury to domestic commerce through an anticompetitive effect or by
making possible anticompetitive acts directed at domestic commerce. In addition, the court
noted that the FTAIA itself also limited the reach of the antitrust laws by requiring that the
“effect” of the conduct be “direct, substantial, and reasonably foreseeable,”83 stating that this
standard would prevent conduct with a merely “ancillary effect” upon U.S. markets from being
actionable under U.S. antitrust laws.
At this point, we have many questions and few answers regarding the Second Circuit’s ruling.
There are significant jurisprudential and policy issues raised by this decision. On the legal side,
for example, the Kruman ruling leaves open issues of proof (will they be insurmountable for the
foreign plaintiffs?) and how to measure the remedies. On the policy side, questions about the
costs and benefits of this decision, if it stands, are numerous. For instance, what does the U.S.
have to gain by allowing such lawsuits to proceed? Is it primarily an additional instrument of
deterrence against anticompetitive behavior? How can we gauge whether the benefits to the
degree of market competition in the U.S. as a result of this rule with outweigh costs to the U.S.
legal system? These questions highlight an important avenue for future research.
82 Id. at *39 n.9. 83 15 U.S.C. §6as (quoted id. at *42).
40
VI. Conclusion
In this investigation of the effects of international cartels on developing countries we have
addressed both developing countries as consumers, as well as developing countries as
competitors or co-conspirators. The cases discussed show the potential for an international cartel
made up of producers from industrialized countries to have simultaneously harmful effects on
developing country consumers and harmful or beneficial effects on developing country
producers.
The multi-dimensional role that U.S. and EU governments play in responding to these cartels has
varied effects on developing countries. The vigorous prosecution of international cartels by the
U.S. and EC may well open up entry possibilities to developing country producers. On the other
hand, these governments are also susceptible to manipulation by domestic producers using tariff
barriers and anti-dumping duties to protect the home market, either during or after the
conspiracy. As raised in the previous section, impending legal decisions on whether to take a
narrow or broad view of the application of U.S. antitrust law will have a direct effect on damages
suffered by developing countries from international cartels.
Finally, there is the role of the antitrust authorities in holding confidential certain information
that could clarify the effect of cartels on developing country consumers and producers. In
general, although U.S. and EC decisions often mention that a cartel had effects “in the U.S. and
elsewhere” or in “certain third markets,” those effects are never included in calculating
punishments. Details regarding the effects of the cartels outside of U.S. and EU markets will, in
general, never be made public. This points to an important weakness in international
competition policy as it affects developing countries. The competition authorities may well have
information regarding restrictions on competition in developing countries, but under current law
and agreements there is often not permission, let alone responsibility, to share that information
with the affected parties. These and other issues will need to be addressed as international
anticompetitive conduct collides with national competition policies.
41
TA
BL
E 1
CIT
RIC
AC
ID:
CA
PAC
ITY
AN
D M
AR
KE
T S
HA
RE
S O
F M
AJO
R F
IRM
S
Firm
Nam
e C
apac
ity –
199
0a (m
illio
n lb
s/yr
) C
apac
ity –
199
5a (m
illio
n lb
s/yr
) U
.S. C
apac
ity –
19
99b (m
illio
n lb
s/yr
)
Glo
bal C
arte
l M
arke
t Sha
re –
19
91c
Glo
bal C
arte
l M
arke
t Sha
re –
19
95c
Pfiz
er/A
DM
d 14
0 18
0
AD
M
200
12%
8%
B
ayer
/Haa
rman
n &
R
eim
er
140
150
24
%
14%
Car
gill
55
160
165
A.E
. Sta
ley
(a
divi
sion
of T
ate
&
Lyle
)
135
Jung
bunz
laue
r 13
2 46
3
11%
21
%
Bay
er
90
160
H
offm
ann-
La
Roc
he
77e
154
13
%
7%
Bio
cor (
Italy
) 53
88
Palc
itric
(Ita
ly)
0 77
Citu
rgia
B
ioch
emic
als
Chi
na (v
ario
us
com
pani
es)
186
535
Akt
iva
0 66
God
ot Is
rael
20
e 40
e
Wor
ld C
apac
ity,
exce
pt fo
r for
mer
U
SSR
877
2,23
0
Wor
ld
Con
sum
ptio
ne 85
0 1,
560
42
Firm
Nam
e C
apac
ity –
199
0a (m
illio
n lb
s/yr
) C
apac
ity –
199
5a (m
illio
n lb
s/yr
) U
.S. C
apac
ity –
19
99b (m
illio
n lb
s/yr
)
Glo
bal C
arte
l M
arke
t Sha
re –
19
91c
Glo
bal C
arte
l M
arke
t Sha
re –
19
95c
Euro
pe
cons
umpt
ione
339
555
U.S
. con
sum
ptio
n es
timat
ese
300
475
U.S
. exp
orts
e 17
52
U.S
. im
ports
65
12
5
TO
TAL
CA
RTE
L M
AR
KET
SH
AR
E 60
%
50%
Not
es fo
r Tab
le 1
: a Th
e nu
mbe
rs in
the
first
two
capa
city
col
umns
are
from
Con
nor (
1998
), Ta
ble
1. T
here
are
alw
ays d
iffer
ence
s in
estim
ates
of c
apac
ity.
For
exam
ple,
a 1
994
Che
mic
al M
arke
ting
Repo
rter
arti
cle
"Citr
ic A
cid
Mar
ket G
row
s with
‘New
Age
’ Drin
k Sa
les"
con
tain
s diff
eren
t est
imat
es fo
r U
.S. p
rodu
ctio
n ca
paci
ty:
AD
M is
repo
rted
at 1
80 m
illio
n po
unds
inst
ead
of C
onno
r's 1
40, H
aarm
ann
& R
eim
er a
t 150
(sam
e as
Con
nor)
, and
C
argi
ll at
130
inst
ead
of C
onno
r's 1
60.
b Pur
chas
ing
Mag
azin
e Se
ptem
ber 7
, 200
0.
c Glo
bal m
arke
t sha
re b
ased
on
capa
city
: C
onno
r (19
98),
Tabl
e 4.
d A
DM
acq
uire
d Pf
izer
’s N
orth
Car
olin
a pl
ant i
n D
ecem
ber 1
990,
and
Pfiz
er c
ontin
ued
to su
pply
citr
ic a
cid
from
its G
roto
n pl
ant u
ntil
mid
-199
3 w
hen
the
plan
t was
clo
sed:
Con
nor (
1998
), p.
22.
e E
stim
ate
by C
onno
r (19
98),
p. 2
3.
43
TA
BL
E 2
CA
RT
EL
ME
MB
ER
S A
ND
GO
VE
RN
ME
NT
FIN
ES:
CIT
RIC
AC
ID A
ND
GR
APH
ITE
EL
EC
TR
OD
ES
C
ITR
IC A
CID
GR
APH
ITE
EL
EC
TR
OD
ES
Firm
Nam
e Fi
ne
Fi
rm N
ame
Fine
A
DM
(U.S
.) $3
0m (D
OJ)
$2
m (C
anad
a)
€36.
7m (E
U)
U
CA
R (U
.S.)
$110
m (D
OJ)
$1
1m (C
anad
a)
€50.
4m (E
U)
Bay
er/H
aarm
ann
& R
eim
er
(Ger
man
y/U
.S.)
$50m
(DO
J)
$4.7
m (C
anad
a)
€14.
22m
(EU
)
SG
L (G
erm
any)
$1
35m
(DO
J)
$12.
5m (C
anad
a)
€80.
2m (E
U)
Jung
bunz
laue
r Int
l. A
G
(Sw
itzer
land
) $1
1m (D
OJ)
$1
.9m
(Can
ada)
€1
7.64
m (E
U)
C
arbi
de/G
raph
ite G
roup
(U.S
.) G
rant
ed le
nien
cy (D
OJ)
€1
0.3m
(EU
)
Hof
fman
n-La
Roc
he
(Sw
itzer
land
) $1
4m (D
OJ)
€6
3.5m
(EU
)
Show
a D
enko
(Jap
an)
$32.
5m (D
OJ)
€1
7.4m
(EU
) C
eres
tar B
iopr
oduc
ts B
V
(Dut
ch su
bsid
iary
of t
he
Fren
ch a
gric
ultu
ral
prod
ucts
firm
Erid
ania
B
eghi
n-Sa
y SA
)
$400
,000
(DO
J)
€0.1
7m (E
U)
To
kai C
arbo
n (J
apan
) $6
m (D
OJ)
€2
4.5m
(EU
)
SE
C C
orp
(Jap
an)
$4.8
m (D
OJ)
€1
2.2m
(EU
)
Nip
pon
Car
bon
(Jap
an)
$2.5
m (D
OJ)
€1
2.2m
(EU
)
VA
W A
lum
inum
(Ger
man
y)
€11.
6m (E
U)
44
TA
BL
E 3
GR
APH
ITE
EL
EC
TR
OD
ES:
CA
PAC
ITY
AN
D M
AR
KE
T S
HA
RE
S O
F M
AJO
R F
IRM
S
Firm
Nam
e U
.S. M
arke
t Sha
re
(at t
ime
of c
onsp
irac
y,
as r
epor
ted
by D
OJ)
Wor
ld M
arke
t Sha
re
(199
9, a
s sta
ted
in
Ferr
omin
com
plai
nt)
UC
AR
34
%
31%
SG
L 23
%
27%
C
arbi
de/G
raph
ite
Gro
up
18%
6%
Show
a D
enko
18
%
6%
Toka
i Car
bon
1%
11%
SE
C C
orp
5%
N
ippo
n C
arbo
n
4%
45
TA
BL
E 4
GL
OB
AL
EL
EC
TR
IC A
RC
FU
RN
AC
E P
RO
DU
CT
ION
IN 2
000
R
egio
n Pr
oduc
tion
(mill
ion
met
ric
tons
)
Oxy
gen
%
Ele
ctri
c
%
Ope
n H
eart
h
%
EU
163.
2 60
.3
39.7
0
Oth
er E
urop
e 46
.5
56.8
41
.7
1.6
Form
er U
SSR
98
.6
54.6
12
.5
32.9
N
AFT
A
133.
8 51
.5
48.5
Cen
tral &
Sou
th
Am
eric
a 40
.3
65.0
35
.0
Afr
ica
13.4
49
.3
50.1
Mid
dle
East
10
.6
20.8
79
.2
A
sia
330.
3 62
.1
28.5
1.
7 W
ORL
D
845.
8 58
.6
33.8
4.
6 So
urce
: In
tern
atio
nal I
ron
and
Stee
l Ins
titut
e, T
rend
s and
Indi
cato
rs: W
orld
Ste
el P
rodu
ctio
n D
ata,
at h
ttp://
ww
w.w
orld
stee
l.org
/tren
ds_p
rod/
prod
06 (v
isite
d N
ovem
ber 2
000)
.
46
TA
BL
E 5
OC
TG
: M
AJO
R W
OR
LD
WID
E P
RO
DU
CE
RS
Cor
pora
te N
ame
Loc
atio
n(s)
of
OC
TG
pro
duct
ion
Car
tel
Part
icip
ant
Gru
po D
ST (T
echi
nt)
NK
K (a
llied
with
DST
) Ja
pan
Yes
D
alm
ine
(par
t of D
ST g
roup
)
Form
erly
ow
ned
by Il
va S
pA
Italy
Y
es
Side
rca
(par
t of D
ST g
roup
) A
rgen
tina
No
Tam
sa (p
art o
f DST
gro
up)
Mex
ico
No
NK
KTu
bes (
join
t ven
ture
of N
KK
and
Si
derc
a)
Alg
oma
Seam
less
Tub
ular
s C
anad
a N
o
Vallo
urec
-Man
nesm
ann
Allia
nce
Man
nesm
ann
AG
M
anne
sman
n R
ohre
n-W
erke
G
erm
any
Yes
M
anne
sman
n H
ande
l (no
w o
wne
d by
Th
ysse
n)
Ger
man
y In
vest
igat
ed,
but n
ot fi
ned
Man
nesm
ann
SA B
razi
l B
razi
l
Val
lour
ec
Fran
ce
Yes
V
allo
urec
& M
anne
sman
n Tu
bes
(V&
M T
ubes
or V
&M
do
Bra
sil,
join
t ve
ntur
e of
Man
nesm
ann
and
Val
lour
ec)
Bra
zil
No
47
Cor
pora
te N
ame
Loc
atio
n(s)
of
OC
TG
pro
duct
ion
Car
tel
Part
icip
ant
Euro
pipe
(joi
ntly
ow
ned
by
Man
nesm
ann,
Usi
nor,
and
Cor
us)
Ger
man
y, F
ranc
e,
US
Inve
stig
ated
, bu
t not
fine
d Th
ysse
n St
ahlu
nion
(ow
ns sm
all s
hare
of
Man
nesm
ann
Roh
ren-
Wer
ke)
Ger
man
y In
vest
igat
ed,
but n
ot fi
ned
“Jap
anes
e” A
llian
ce
Kaw
asak
i Ja
pan
Yes
N
ippo
n St
eel
Japa
n Y
es
Sum
itom
o M
etal
(Sum
itom
o D
euts
chla
nd)
Japa
n Y
es
O
ther
Pro
duce
rs
Sand
vik
Stee
l Sw
eden
, US,
C
anad
a, C
zech
R
epub
lic, F
ranc
e,
Ger
man
y, In
dia
No
Cor
us (f
orm
erly
Brit
ish
Stee
l) U
K
Yes
48
T
AB
LE
6
R
EC
EN
T IN
TE
RN
AT
ION
AL
CA
RT
EL
S IN
VE
STIG
AT
ED
BY
TH
E U
.S. D
EPA
RT
ME
NT
OF
JUST
ICE
AN
D T
HE
EU
RO
PEA
N C
OM
MIS
SIO
N
In
dust
ry
Star
t1 E
nd
Con
vict
ion
C
ount
ry o
f Ori
gin
of In
dict
ed
Firm
s D
evel
opin
g C
ount
ry
Part
icip
ants
Cou
ntry
(ies)
K
now
n T
o B
e A
ffec
ted2
Alu
min
um P
hosp
hide
Ja
n 19
90
Nov
199
0 D
OJ
Bra
zil,
Ger
man
y, In
dia,
US
Bra
zil,
Indi
a U
S B
eer
1993
19
98
EC
Bel
gium
, Fra
nce
No
Bel
gium
B
rom
ine
Prod
ucts
Ju
l 199
5 A
pr 1
998
DO
J Is
rael
, US
No
US
Cab
le-S
taye
d B
ridge
s Se
p 19
96
Dec
199
7 D
OJ
Fran
ce, U
S N
o U
S C
arbo
n C
atho
de B
lock
Fe
b 19
96
Dec
199
7 D
OJ
Ger
man
y, Ja
pan,
US
No
US
and
else
whe
re
Car
bonl
ess P
aper
19
92
1995
EC
Fr
ance
, Ger
man
y, S
pain
, Sou
th
Afr
ica,
UK
So
uth
Afr
ica
Euro
pe
Car
tonb
oard
19
86
1991
EC
3
Aus
tria,
Can
ada,
Fin
land
, Fr
ance
, Ger
man
y, It
aly,
N
ethe
rland
s, N
orw
ay, S
pain
, Sw
eden
, UK
, US
(via
Eur
opea
n su
bsid
iarie
s)
No
Euro
pe
Cem
ent
1983
A
ug 1
994
EC
33 E
urop
ean
firm
s, 8
natio
nal
cem
ent t
rade
ass
ocia
tions
, and
th
e Eu
rope
an C
emen
t A
ssoc
iatio
n
No
Euro
pe
Citr
ic A
cid
1991
19
95
DO
J & E
C
Aus
tria,
Fra
nce,
Ger
man
y,
Net
herla
nds,
Switz
erla
nd, U
S N
o In
tern
atio
nal
Expl
osiv
es
1988
19
92
DO
J N
orw
ay, U
K, U
S N
o U
S Fe
rros
ilico
n O
ct 1
989
June
199
1 D
OJ
Ger
man
y, N
orw
ay, U
S N
o In
tern
atio
nal
Ferr
y O
pera
tors
(Adr
iatic
Se
a)
1987
19
94
EC
Gre
ece,
Ital
y N
o G
reec
e, It
aly
Ferr
y O
pera
tors
(Cro
ss-
Cha
nnel
Fre
ight
) O
ct 1
992
Dec
199
2 EC
Fr
ance
, Net
herla
nds,
Swed
en,
UK
N
o Eu
rope
49
Indu
stry
St
art1
End
C
onvi
ctio
n
Cou
ntry
of O
rigi
n of
Indi
cted
Fi
rms
Dev
elop
ing
Cou
ntry
Pa
rtic
ipan
ts
Cou
ntry
(ies)
K
now
n T
o B
e A
ffec
ted2
Fine
Arts
A
pr 1
993
Dec
199
9 D
OJ (
EC is
sued
“S
tate
men
t of
Obj
ectio
ns”
on
Apr
il 19
, 200
2)
UK
, US
No
Aus
tralia
, Jap
an,
Uni
ted
Stat
es,
Euro
pe
Gra
phite
Ele
ctro
des
Jul 1
992
Jun
1997
D
OJ &
EC
G
erm
any,
Japa
n, U
S N
o In
tern
atio
nal
Isos
tatic
Gra
phite
Ju
l 199
3 Fe
b 19
98
DO
J Ja
pan,
US
No
Can
ada,
US
La
min
ated
Pla
stic
Tub
es
1987
19
96
DO
J Sw
itzer
land
, US
No
US
Lysi
ne
Jun
1992
Ju
n 19
95
DO
J & E
C
Ger
man
y, Ja
pan,
Sou
th K
orea
, U
S So
uth
Kor
ea
Inte
rnat
iona
l
Mal
tol
Dec
198
9 D
ec 1
995
DO
J U
S +
unna
med
firm
s N
o U
S an
d el
sew
here
M
arin
e C
onst
ruct
ion
Serv
ices
(Hea
vy-L
ift)
1993
19
97
DO
J N
ethe
rland
s, U
S
No
US
and
else
whe
re
Mar
ine
Tran
spor
tatio
n Se
rvic
es (H
eavy
-Lift
) 19
90
1995
D
OJ
Bel
gium
, US
No
US
and
else
whe
re
Mon
ochl
oroa
cetic
Aci
d Se
p 19
95
Aug
199
9 D
OJ
Fran
ce, N
ethe
rland
s N
o N
A
Nuc
leot
ides
Ju
l 199
2 A
ug 1
996
DO
J Ja
pan,
Sou
th K
orea
So
uth
Kor
ea
NA
O
rgan
ic P
erox
ides
19
97
1998
D
OJ
Fran
ce +
unn
amed
firm
s N
A
NA
Pl
astic
Din
nerw
are
Nov
199
1 A
pr 1
992
DO
J C
anad
a, U
S N
o U
S Sh
ippi
ng (C
entra
l Wes
t A
fric
an)
1972
19
92
EC (c
onvi
ctio
n,
but f
ine
over
turn
ed)
Zaire
, Ang
ola,
Nor
ther
n pa
rt of
co
ntin
enta
l Eur
ope,
exc
ludi
ng
UK
Zaire
, Ang
ola
NA
Ship
ping
(Far
Eas
t) 19
91
1994
EC
30
cou
ntrie
s M
alay
sia,
So
uth
Kor
ea
Inte
rnat
iona
l
Ship
ping
(Fre
nch-
Afr
ican
) 19
75
1992
EC
12
cou
ntrie
s e.
g., S
eneg
al,
Cam
eroo
n Fr
ance
, Sen
egal
, G
abon
, Cen
tral
Afr
ican
Rep
ublic
, N
iger
, Bur
kina
, Fa
so, G
uine
a,
Con
go, M
ali,
Togo
, an
d C
amer
oon
Ship
ping
(Nor
th A
tlant
ic)
1994
19
96
EC
11 c
ount
ries
e.g.
, Pol
and
Inte
rnat
iona
l So
dium
Ery
thor
bate
Ju
l 199
2 D
ec 1
994
DO
J U
S +
unna
med
firm
s N
o U
S
50
Indu
stry
St
art1
End
C
onvi
ctio
n
Cou
ntry
of O
rigi
n of
Indi
cted
Fi
rms
Dev
elop
ing
Cou
ntry
Pa
rtic
ipan
ts
Cou
ntry
(ies)
K
now
n T
o B
e A
ffec
ted2
Sodi
um G
luco
nate
A
ug 1
993
June
199
5 D
OJ &
EC
Fr
ance
, Jap
an, N
ethe
rland
s, Sw
itzer
land
, US
No
Inte
rnat
iona
l
Sorb
ates
19
79
1996
D
OJ
Ger
man
y, Ja
pan,
US
No
Inte
rnat
iona
l St
ainl
ess S
teel
Ja
n 19
94
Mar
199
5 EC
B
elgi
um, F
ranc
e, G
erm
any,
Ita
ly, S
pain
, Sw
eden
, UK
N
o Eu
rope
Stee
l Bea
m
1988
19
94
EC
Bel
gium
, Fra
nce,
Ger
man
y,
Luxe
mbo
urg,
Spa
in, U
K
No
W. E
urop
e
Stee
l Hea
ting
Pipe
(P
re-I
nsul
ated
Pip
e)
late
199
0 19
96
EC
Aus
tria,
Den
mar
k, F
inla
nd,
Ger
man
y, It
aly,
Sw
eden
, Sw
itzer
land
No
Euro
pe
Stee
l Tub
e, S
eam
less
19
90
1995
EC
Fr
ance
, Ger
man
y, It
aly,
Japa
n,
UK
N
o Eu
rope
and
“ce
rtain
th
ird m
arke
ts”
Suga
r Ju
n 19
86
Jul 1
990
EC
Den
mar
k, Ir
elan
d, U
K
No
UK
Ta
mpi
co F
iber
Ja
n 19
90
Apr
199
5 D
OJ
Mex
ico,
Net
herla
nds,
US
Mex
ico
US
Ther
mal
Fax
Pap
er
1991
19
92
DO
J (1
Japa
nese
fir
m w
ent t
o tri
al
and
won
)
Japa
n, U
S N
o N
orth
Am
eric
a
Vita
min
s Ja
n 19
90
Feb
1999
D
OJ &
EC
C
anad
a, G
erm
any,
Japa
n,
Switz
erla
nd, U
S N
o In
tern
atio
nal
Was
tew
ater
Con
stru
ctio
n Ju
n 19
88
Jan
1995
D
OJ
Ger
man
y, S
witz
erla
nd, U
S N
o Eg
ypt
Zinc
Pho
spha
te
Mar
199
4 M
ay 1
998
EC
Fran
ce, G
erm
any,
Nor
way
, UK
N
o Eu
rope
N
otes
to T
able
6:
1.
Car
tel d
ates
are
app
roxi
mat
e. I
n pa
rticu
lar,
indi
ctm
ents
of d
iffer
ent f
irms m
ay li
st d
iffer
ent c
onsp
iracy
dat
es.
Als
o, fo
r tho
se c
arte
ls
pros
ecut
ed b
y bo
th th
e D
OJ a
nd E
C, w
e lis
t the
DO
J dat
es o
nly;
som
etim
es th
e EC
cas
es w
ill h
ave
diff
eren
t dat
es.
In g
ener
al, t
he in
form
atio
n pr
esen
ted
in th
is ta
ble
was
gat
here
d fr
om v
ario
us in
dust
ry a
nd g
over
nmen
t sou
rces
, inc
ludi
ng D
OJ a
nd E
C p
ress
rele
ases
, Eur
opea
n C
ourt
of
Just
ice
deci
sion
s, an
d in
dust
ry a
nd b
usin
ess n
ews s
ourc
es, s
uch
as A
mer
ican
Met
al M
arke
t, C
hem
ical
Mar
ketin
g Re
port
er, E
urop
ean
Busi
ness
W
eek,
Inte
rnat
iona
l Cem
ent M
agaz
ine,
Oil
and
Gas
Jou
rnal
, and
Wal
l Str
eet J
ourn
al.
Spec
ific
sour
ces a
re a
vaila
ble
from
the
auth
ors u
pon
requ
est.
2.
Info
rmat
ion
on "
Cou
ntry
(ies)
Kno
wn
To B
e A
ffec
ted"
repo
rted
in th
is ta
ble
com
es fr
om D
OJ a
nd E
C p
ress
rele
ases
, ind
ictm
ents
, and
rulin
gs,
as w
ell a
s arti
cles
in th
e pr
ess.
The
se d
ocum
ents
, of c
ours
e, fo
cus o
n th
e ef
fect
s in
eith
er th
e U
nite
d St
ates
or E
urop
e. I
n m
ost c
ases
ther
e is
no
info
rmat
ion
from
thes
e so
urce
s on
who
pur
chas
ed fr
om th
e ca
rtel.
3.
Com
pani
es a
ppea
led,
but
Cou
rt of
Firs
t Ins
tanc
e co
nfirm
ed th
e ba
sic
deci
sion
, alth
ough
ann
ullin
g m
inor
par
ts o
f the
dec
isio
n.
51
T
AB
LE
7
RE
CE
NT
INT
ER
NA
TIO
NA
L C
AR
TE
LS:
R
EPO
RT
ED
MA
RK
ET
CO
NC
EN
TR
AT
ION
AN
D P
RIC
E IN
CR
EA
SES
Indu
stry
Mea
sure
of M
arke
t Co
ncen
tratio
na M
easu
re o
f Pric
e In
crea
seb
Alu
min
um P
hosp
hide
U
.S.:
C4
= 90
48
%
Bro
min
e Pr
oduc
ts
Wor
ld: C
3 =
76
NA
Car
bonl
ess P
aper
Eu
rope
: C
arte
l sha
re =
85-
90
NA
Car
tonb
oard
Eu
rope
: C
arte
l sha
re =
80
20-2
6%
Cem
ent
Euro
pe: C
6 =
50
NA
Citr
ic A
cid
Wor
ld: C
4 =
60
W. E
urop
e: C
4= 8
5 Es
timat
ed 2
1-24
% m
arku
p ov
er m
argi
nal c
ost
Rep
orte
d pr
ice
incr
ease
of 3
0% in
the
U.S
. and
50
% in
Eur
ope
Ferr
osili
con
U
.S.:
C4
= 10
0 N
A
Ferr
y O
pera
tors
(C
ross
-Cha
nnel
Fre
ight
)C
2 =
72 fo
r cro
ss-c
hann
el
pass
enge
r mar
ket;
sam
e tw
o fir
ms w
ere
indi
cted
in th
e cr
oss-
chan
nel f
reig
ht c
arte
l
10%
Fine
Arts
C
2 =
95
0 –
20%
Gra
phite
Ele
ctro
des
U.S
.: C
4 =
93
Can
ada:
C2
= 90
U
.S.:
50%
- 60
%
Can
ada:
90%
52
Indu
stry
Mea
sure
of M
arke
t Co
ncen
tratio
na M
easu
re o
f Pric
e In
crea
seb
Isos
tatic
Gra
phite
W
orld
: 6
maj
or fi
rms
NA
Lam
inat
ed P
last
ic T
ubes
U.S
.: C
3 =
95
NA
Lysi
ne
Wor
ld: C
3 =
95 in
late
198
0’s
(AD
M b
egan
pro
duct
ion
in 1
991
and
by 1
996
had
47%
wor
ld
mar
ket s
hare
) M
exic
o: C
2 =
90
Rep
orte
d pr
ice
incr
ease
s of 4
1% (“
wor
ld”)
, 50%
(C
anad
a), a
nd 7
0% in
the
first
six
mon
ths,
doub
ling
over
the
life
of th
e co
nspi
racy
(U.S
.)
Mal
tol
Wor
ld: C
2 =
80-9
0 N
A
Plas
tic D
inne
rwar
e U
.S.:
C2
> 90
N
A
Ship
ping
(Cen
tral W
est
Afr
ican
) Sh
ippi
ng c
onfe
renc
e he
ld m
ore
than
90%
mar
ket s
hare
N
A
Ship
ping
(Far
Eas
t) Sh
ippi
ng c
onfe
renc
e he
ld 8
0%
shar
e be
twee
n no
rther
n Eu
rope
an
d th
e Fa
r Eas
t
NA
Ship
ping
(Fre
nch-
Afr
ican
) C
4 =
90
34-3
9%
Ship
ping
(Nor
th
Atla
ntic
) 19
94: C
arte
l mem
bers
had
join
t m
arke
t sha
re o
ver 6
0%
1995
: Shi
ppin
g co
nfer
ence
hel
d be
twee
n 70
-80%
of t
he tr
ans
Nor
th-A
tlant
ic c
onta
iner
mar
ket
Dam
ages
Est
imat
e, 1
995:
Ann
ounc
ed p
rice
incr
ease
s in
1995
“w
ould
hav
e m
eant
an
addi
tiona
l cos
t to
ship
pers
of U
S$65
to $
75
mill
ion,
whe
n co
mpa
red
with
199
4 pr
ices
” D
amag
es E
stim
ate,
199
4-98
: Tra
ns-A
tlant
ic
Con
fere
nce
Agr
eem
ent (
TAC
A) c
ost E
urop
ean
indu
stry
$1.
8 bi
llion
in e
xtra
ship
ping
cos
ts
betw
een
1994
and
199
8
Sodi
um E
ryth
orba
te
U.S
.: C
2 =
90
NA
53
Indu
stry
Mea
sure
of M
arke
t Co
ncen
tratio
na M
easu
re o
f Pric
e In
crea
seb
Sodi
um G
luco
nate
C
arte
l mem
bers
wer
e w
orld
’s
maj
or p
rodu
cers
N
A
Sorb
ates
W
orld
: C2
= 71
14
%
Stai
nles
s Ste
el
Wor
ld: C
4 =
52
100%
Stee
l Bea
m
Euro
pe:
C10
= 6
6 N
A
Stee
l Hea
ting
Pipe
(Pre
-In
sula
ted
Pipe
) W
este
rn E
urop
e: C
4 =
80
15-2
0% o
ver o
ne y
ear
Stee
l Tub
e, S
eam
less
Eu
rope
car
tel s
hare
= 1
9 N
A
Suga
r G
reat
Brit
ain:
C2
= 90
N
A
Tam
pico
Fib
er
Car
tel m
embe
rs h
ad
“ove
rwhe
lmin
g” sh
are
of th
e U
.S. m
arke
t
NA
Ther
mal
Fax
Pap
er
U.S
.: C
3 =
40-4
5 10
%
Vita
min
s W
orld
: C3
= 75
for b
ulk
vita
min
s U
.S.:
20%
C
anad
a: 3
0%
Zinc
Pho
spha
te
Euro
pe c
arte
l sha
re =
90
NA
54
Not
es to
Tab
le 7
:
a A
ll co
ncen
tratio
n fig
ures
are
app
roxi
mat
e. W
here
ver p
ossi
ble
conc
entra
tion
mea
sure
s dat
e to
the
perio
d of
the
carte
l. In
oth
er
inst
ance
s the
figu
res d
ate
to th
e pe
riod
imm
edia
tely
prio
r to
or a
fter t
he c
arte
l. D
etai
led
refe
renc
es a
re a
vaila
ble
from
the
auth
ors.
b A
ll pr
ice
incr
ease
mea
sure
s are
app
roxi
mat
e an
d co
me
from
the
trade
pre
ss a
nd p
ress
rele
ases
, off
icia
l dec
isio
ns, a
nd sp
eech
es m
ade
by c
ompe
titio
n po
licy
auth
oriti
es.
Info
rmat
ion
is e
xtre
mel
y sp
arse
. Th
e pr
ice
incr
ease
info
rmat
ion
usua
lly re
fers
to a
sele
cted
per
iod
with
in th
e ca
rtel y
ears
; it i
s not
nor
mal
ly a
n in
dica
tor o
f the
ave
rage
pric
e in
crea
se o
ver t
he e
ntire
life
of t
he c
arte
l. D
etai
led
refe
renc
es a
re a
vaila
ble
from
the
auth
ors.
55
TA
BL
E 8
IMPO
RT
AN
CE
OF
CA
RT
EL
IZE
D G
OO
DS
IN D
EV
EL
OPI
NG
CO
UN
TR
Y IM
POR
TS,
199
71
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Alu
min
um P
hosp
hide
2 0.
0078
%
0.00
48%
0.
0074
%
0.00
67%
0.
0065
%
Bee
r3 0.
0091
%
0.01
47%
0.
0072
%
0.00
98%
0.
0182
%
Bro
min
e4 0.
0166
%
0.02
84%
0.
0346
%
0.02
98%
0.
0300
%
Cab
le-S
taye
d B
ridge
s5 0.
0796
%
0.07
38%
0.
0495
%
0.06
18%
0.
0398
%
Car
bon
Cat
hode
Blo
ck6
0.06
86%
0.
2232
%
0.25
72%
0.
2163
%
0.00
00%
Car
bonl
ess P
aper
7 0.
0158
%
0.02
48%
0.
0234
%
0.02
26%
0.
0210
%
Car
tonb
oard
8 0.
0200
%
0.03
45%
0.
0502
%
0.04
05%
0.
0250
%
Cem
ent9
0.05
59%
0.
0258
%
0.02
49%
0.
0302
%
0.01
34%
Citr
ic A
cid10
0.
1183
%
0.10
85%
0.
0778
%
0.09
38%
0.
0528
%
Expl
osiv
es11
N
A
NA
N
A
NA
N
A
Ferr
osili
con12
0.
0122
%
0.00
73%
0.
0226
%
0.01
62%
0.
0230
%
Fine
Arts
13
0.00
28%
0.
0050
%
0.00
78%
0.
0061
%
0.03
42%
Gra
phite
Ele
ctro
des14
0.
0686
%
0.22
32%
0.
2572
%
0.21
63%
0.
0000
%
Isos
tatic
Gra
phite
15
0.14
84%
0.
2265
%
0.22
52%
0.
2132
%
0.14
01%
Lam
inat
ed P
last
ic T
ubes
16
0.36
89%
1.
1664
%
0.69
06%
0.
7852
%
0.47
74%
Lysi
ne17
0.
1113
%
0.09
32%
0.
1230
%
0.11
19%
0.
1102
%
Mal
tol18
0.
0623
%
0.07
78%
0.
1112
%
0.09
30%
0.
0900
%
Mar
ine
Con
stru
ctio
n19
0.21
91%
0.
0919
%
0.33
72%
0.
2426
%
0.07
62%
Mon
ochl
oroa
cetic
Aci
d20
0.11
83%
0.
1085
%
0.07
78%
0.
0938
%
0.05
28%
56
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Nuc
leot
ides
21
NA
N
A
NA
N
A
NA
O
rgan
ic P
erox
ides
22
0.11
83%
0.
1085
%
0.07
78%
0.
0938
%
0.05
28%
Plas
tic D
inne
rwar
e23
0.05
44%
0.
1461
%
0.21
15%
0.
1660
%
0.13
77%
Sodi
um E
ryth
orba
te24
0.
0654
%
0.07
87%
0.
0586
%
0.06
59%
0.
0228
%
Sodi
um G
luco
nate
25
0.06
23%
0.
0778
%
0.11
12%
0.
0930
%
0.09
00%
Sorb
ates
26
0.11
83%
0.
1085
%
0.07
78%
0.
0938
%
0.05
28%
Stai
nles
s Ste
el27
0.
1183
%
0.10
85%
0.
0778
%
0.09
38%
0.
0528
%
Stee
l Bea
m28
0.
2605
%
0.60
74%
0.
3308
%
0.40
46%
0.
2011
%
Stee
l Hea
ting
Pipe
29
0.06
54%
0.
0787
%
0.05
86%
0.
0659
%
0.02
28%
Stee
l Tub
e, S
eam
less
30
0.02
62%
0.
0416
%
0.03
43%
0.
0353
%
0.02
25%
Suga
r31
0.12
65%
0.
0718
%
0.01
38%
0.
0498
%
0.01
09%
Tam
pico
Fib
er32
0.
0006
%
0.00
25%
0.
0014
%
0.00
16%
0.
0014
%
Ther
mal
Fax
Pap
er33
0.
0611
%
0.12
94%
0.
0761
%
0.09
01%
0.
0783
%
Vita
min
s34
0.01
00%
0.
0212
%
0.01
49%
0.
0161
%
0.01
30%
Was
tew
ater
Con
stru
ctio
n35
0.05
61%
0.
1051
%
0.09
16%
0.
0900
%
0.04
67%
Zinc
Pho
spha
te36
N
A
NA
N
A
NA
N
A
Tota
l 12
.96%
12
.60%
11
.77%
12
.19%
8.
97%
Not
es to
Tab
les 8
, 9, a
nd 1
0 ar
e gi
ven
at th
e en
d of
Tab
le 1
0.
57
TA
BL
E 9
CA
RT
EL
IZE
D G
OO
DS
AS
A S
HA
RE
OF
DE
VE
LO
PIN
G C
OU
NT
RY
GD
P, 1
9971
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Perc
ent o
f GD
P Pe
rcen
t of G
DP
Perc
ent o
f GD
P Pe
rcen
t of G
DP
Perc
ent o
f GD
P
Alu
min
um P
hosp
hide
2 0.
0078
%
0.00
48%
0.
0074
%
0.00
67%
0.
0065
%
Bee
r3 0.
0091
%
0.01
47%
0.
0072
%
0.00
98%
0.
0182
%
Bro
min
e4 0.
0166
%
0.02
84%
0.
0346
%
0.02
98%
0.
0300
%
Cab
le-S
taye
d B
ridge
s5 0.
0796
%
0.07
38%
0.
0495
%
0.06
18%
0.
0398
%
Car
bon
Cat
hode
Blo
ck6
0.06
86%
0.
2232
%
0.25
72%
0.
2163
%
0.00
00%
Car
bonl
ess P
aper
7 0.
0158
%
0.02
48%
0.
0234
%
0.02
26%
0.
0210
%
Car
tonb
oard
8 0.
0200
%
0.03
45%
0.
0502
%
0.04
05%
0.
0250
%
Cem
ent9
0.05
59%
0.
0258
%
0.02
49%
0.
0302
%
0.01
34%
Citr
ic A
cid10
0.
1183
%
0.10
85%
0.
0778
%
0.09
38%
0.
0528
%
Expl
osiv
es11
N
A
NA
N
A
NA
N
A
Ferr
osili
con12
0.
0122
%
0.00
73%
0.
0226
%
0.01
62%
0.
0230
%
Fine
Arts
13
0.00
28%
0.
0050
%
0.00
78%
0.
0061
%
0.03
42%
Gra
phite
Ele
ctro
des14
0.
0686
%
0.22
32%
0.
2572
%
0.21
63%
0.
0000
%
Isos
tatic
Gra
phite
15
0.14
84%
0.
2265
%
0.22
52%
0.
2132
%
0.14
01%
Lam
inat
ed P
last
ic T
ubes
16
0.36
89%
1.
1664
%
0.69
06%
0.
7852
%
0.47
74%
Lysi
ne17
0.
1113
%
0.09
32%
0.
1230
%
0.11
19%
0.
1102
%
Mal
tol18
0.
0623
%
0.07
78%
0.
1112
%
0.09
30%
0.
0900
%
Mar
ine
Con
stru
ctio
n19
0.21
91%
0.
0919
%
0.33
72%
0.
2426
%
0.07
62%
Mon
ochl
oroa
cetic
Aci
d20
0.11
83%
0.
1085
%
0.07
78%
0.
0938
%
0.05
28%
58
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Perc
ent o
f GD
P Pe
rcen
t of G
DP
Perc
ent o
f GD
P Pe
rcen
t of G
DP
Perc
ent o
f GD
P
Nuc
leot
ides
21
NA
N
A
NA
N
A
NA
Org
anic
Per
oxid
es22
0.
1183
%
0.10
85%
0.
0778
%
0.09
38%
0.
0528
%
Plas
tic D
inne
rwar
e23
0.05
44%
0.
1461
%
0.21
15%
0.
1660
%
0.13
77%
Sodi
um E
ryth
orba
te24
0.
0654
%
0.07
87%
0.
0586
%
0.06
59%
0.
0228
%
Sodi
um G
luco
nate
25
0.06
23%
0.
0778
%
0.11
12%
0.
0930
%
0.09
00%
Sorb
ates
26
0.11
83%
0.
1085
%
0.07
78%
0.
0938
%
0.05
28%
Stai
nles
s Ste
el27
0.
1183
%
0.10
85%
0.
0778
%
0.09
38%
0.
0528
%
Stee
l Bea
m28
0.
2605
%
0.60
74%
0.
3308
%
0.40
46%
0.
2011
%
Stee
l Hea
ting
Pipe
29
0.06
54%
0.
0787
%
0.05
86%
0.
0659
%
0.02
28%
Stee
l Tub
e, S
eam
less
30
0.02
62%
0.
0416
%
0.03
43%
0.
0353
%
0.02
25%
Suga
r31
0.12
65%
0.
0718
%
0.01
38%
0.
0498
%
0.01
09%
Tam
pico
Fib
er32
0.
0006
%
0.00
25%
0.
0014
%
0.00
16%
0.
0014
%
Ther
mal
Fax
Pap
er33
0.
0611
%
0.12
94%
0.
0761
%
0.09
01%
0.
0783
%
Vita
min
s34
0.01
00%
0.
0212
%
0.01
49%
0.
0161
%
0.01
30%
Was
tew
ater
Con
stru
ctio
n35
0.05
61%
0.
1051
%
0.09
16%
0.
0900
%
0.04
67%
Zinc
Pho
spha
te36
N
A
NA
N
A
NA
N
A
Tota
l 1.
97%
3.
32%
2.
71%
2.
78%
1.
69%
Not
es to
Tab
les 8
, 9, a
nd 1
0 ar
e gi
ven
at th
e en
d of
Tab
le 1
0.
59
T
AB
LE
10
TO
TA
L V
AL
UE
OF
CA
RT
EL
-AFF
EC
TE
D IM
POR
TS,
199
71
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Val
ue o
f im
port
s
($00
0)
Val
ue o
f im
port
s
($00
0)
Val
ue o
f im
port
s
($00
0)
Tot
al Im
port
s by
Dev
elop
ing
Cou
ntri
es
($00
0)
Val
ue o
f im
port
s
($00
0)
Alu
min
um P
hosp
hide
2 70
,402
82
,664
22
0,86
0 37
3,92
6 1,
452,
864
Bee
r3 82
,321
25
4,61
2 21
4,36
1 55
1,29
4 4,
053,
841
Bro
min
e4 15
0,14
8 49
1,90
7 1,
031,
786
1,67
3,84
1 6,
673,
615
Cab
le-S
taye
d B
ridge
s5 72
1,14
5 1,
276,
068
1,47
3,14
3 3,
470,
356
8,86
8,03
6 C
arbo
n C
atho
de B
lock
6 62
1,59
5 3,
859,
815
7,65
9,25
2 12
,140
,662
30
,557
,546
C
arbo
nles
s Pap
er7
142,
813
428,
327
696,
821
1,26
7,96
1 4,
682,
645
Car
tonb
oard
8 18
1,25
1 59
5,79
5 1,
495,
838
2,27
2,88
4 5,
576,
247
Cem
ent9
506,
339
445,
984
741,
979
1,69
4,30
2 2,
984,
991
Citr
ic A
cid10
1,
071,
650
1,87
6,18
2 2,
316,
093
5,26
3,92
5 11
,771
,185
Ex
plos
ives
11
NA
N
A
NA
N
A
NA
Fe
rros
ilico
n12
110,
896
126,
282
672,
946
910,
124
5,13
0,07
8 Fi
ne A
rts13
25
,785
86
,150
23
1,26
7 34
3,20
2 7,
619,
113
Gra
phite
Ele
ctro
des14
62
1,59
5 3,
859,
815
7,65
9,25
2 12
,140
,662
30
,557
,546
Is
osta
tic G
raph
ite15
1,
344,
868
3,91
6,39
2 6,
707,
800
11,9
69,0
60
31,2
15,2
17
Lam
inat
ed P
last
ic T
ubes
16
3,34
2,77
8 20
,171
,006
20
,567
,075
44
,080
,859
10
6,36
4,31
4
60
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Val
ue o
f im
port
s
($00
0)
Val
ue o
f im
port
s
($00
0)
Val
ue o
f im
port
s
($00
0)
Tot
al Im
port
s by
Dev
elop
ing
Cou
ntri
es
($00
0)
Val
ue o
f im
port
s
($00
0)
Lysi
ne17
1,
008,
248
1,61
0,87
0 3,
664,
349
6,28
3,46
7 24
,560
,999
M
alto
l18
564,
916
1,34
5,03
5 3,
312,
196
5,22
2,14
7 20
,060
,451
M
arin
e C
onst
ruct
ion19
1,
984,
798
1,58
8,46
1 10
,043
,702
13
,616
,961
16
,978
,097
M
onoc
hlor
oace
tic A
cid20
1,
071,
650
1,87
6,18
2 2,
316,
093
5,26
3,92
5 11
,771
,185
N
ucle
otid
es21
N
A
NA
N
A
NA
N
A
Org
anic
Per
oxid
es22
1,
071,
650
1,87
6,18
2 2,
316,
093
5,26
3,92
5 11
,771
,185
Pl
astic
Din
nerw
are23
49
3,25
5 2,
525,
856
6,29
7,96
0 9,
317,
071
30,6
79,4
51
Sodi
um E
ryth
orba
te24
59
2,19
6 1,
361,
422
1,74
6,71
3 3,
700,
331
5,09
0,47
7 So
dium
Glu
cona
te25
56
4,91
6 1,
345,
035
3,31
2,19
6 5,
222,
147
20,0
60,4
51
Sorb
ates
26
1,07
1,65
0 1,
876,
182
2,31
6,09
3 5,
263,
925
11,7
71,1
85
Stai
nles
s Ste
el27
1,
071,
650
1,87
6,18
2 2,
316,
093
5,26
3,92
5 11
,771
,185
St
eel B
eam
28
2,36
0,02
0 10
,504
,156
9,
850,
996
22,7
15,1
72
44,7
95,2
12
Stee
l Hea
ting
Pipe
29
592,
196
1,36
1,42
2 1,
746,
713
3,70
0,33
1 5,
090,
477
Stee
l Tub
e, S
eam
less
30
236,
976
719,
383
1,02
2,47
7 1,
978,
836
5,01
6,86
9 Su
gar31
1,
146,
447
1,24
0,91
4 41
0,04
8 2,
797,
409
2,42
4,48
4 Ta
mpi
co F
iber
32
5,39
8 43
,522
41
,528
90
,448
30
7,45
9 Th
erm
al F
ax P
aper
33
553,
335
2,23
7,89
5 2,
265,
394
5,05
6,62
4 17
,437
,672
V
itam
ins34
90
,985
36
7,03
4 44
4,80
4 90
2,82
3 2,
887,
216
Was
tew
ater
Con
stru
ctio
n35
508,
359
1,81
6,82
0 2,
727,
370
5,05
2,54
9 10
,406
,321
61
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Val
ue o
f im
port
s
($00
0)
Val
ue o
f im
port
s
($00
0)
Val
ue o
f im
port
s
($00
0)
Tot
al Im
port
s by
Dev
elop
ing
Cou
ntri
es
($00
0)
Val
ue o
f im
port
s
($00
0)
Zinc
Pho
spha
te36
N
A
NA
N
A
NA
N
A
Tota
l 17
,873
,052
57
,379
,791
80
,608
,690
15
5,86
1,53
3 37
7,32
2,78
1 N
otes
to T
able
s 8, 9
, and
10:
1 T
hese
tabl
es u
se a
com
bina
tion
of 3
-dig
it an
d 4-
digi
t SIT
C c
ateg
ory
mat
ches
. Th
e 3-
digi
t cod
es w
ere
used
onl
y w
hen
4-di
git d
ata
wer
e un
avai
labl
e. T
he li
st o
f dev
elop
ing
coun
tries
is ta
ken
from
Wor
ld D
evel
opm
ent R
epor
t 200
0/20
01:
Atta
ckin
g Po
vert
y (W
orld
B
ank)
, pp.
334
-35.
As s
tate
d on
p. 3
35 “
Low
inco
me
and
mid
dle-
inco
me
econ
omie
s are
som
etim
es re
ferr
ed to
as d
evel
opin
g co
untri
es.”
2 In
orga
nic
chem
ical
pro
duct
s, n.
e.s
3 B
eer m
ade
from
mal
t (in
clud
ing
ale,
stou
t and
por
ter)
4 P
heno
ls a
nd p
heny
l alc
ohol
s and
thei
r hal
ogen
ated
der
ivat
ives
; eth
ers,
alco
hol p
erox
ides
, eth
er p
erox
ides
, epo
xide
s etc
; hal
ogen
ated
de
rivat
ives
of h
ydro
carb
ons
5 Stru
ctur
es a
nd p
arts
of s
truct
ures
: iro
n, st
eel,
alum
inum
6 E
lect
rical
mac
hine
ry a
nd a
ppar
atus
, n.e
.s.
7 Pap
er a
nd p
aper
boar
d, c
ut to
size
or s
hape
, n.e
.s.
8 Box
es, b
ags &
oth
er p
acki
ng c
onta
iner
s, of
pap
er o
r pap
erbo
ard
9 Por
tland
cem
ent,
cem
ent f
ondu
, sla
g ce
men
t, et
c.
10 C
arbo
xylic
aci
ds, a
nd th
eir a
nhyd
rides
, hal
ides
, etc
. 11
Exp
losi
ves a
nd p
yrot
echn
ic p
rodu
cts
12 F
erro
-allo
ys
13 A
rt, c
olle
ctor
s’ p
iece
s and
ant
ique
s 14
Oth
er e
lect
rical
mac
hine
ry a
nd e
quip
men
t 15
Org
anic
che
mic
al p
rodu
cts,
n.e.
s.; c
hem
ical
pro
duct
s and
pre
para
tions
, n.e
.s.
16 M
isce
llane
ous a
rticl
es o
f mat
eria
ls o
f div
isio
n 58
; pol
ymer
izat
ion
and
copo
lym
eriz
atio
n pr
oduc
ts; c
onde
nsat
ion,
pol
ycon
dens
atio
n an
d po
lyad
ditio
n pr
oduc
ts
17 N
itrog
en-f
unct
ion
com
poun
ds
62
18 O
rgan
o-in
orga
nic
and
hete
rocy
clic
com
poun
ds
19 S
hips
, boa
ts a
nd fl
oatin
g st
ruct
ures
20
Car
boxy
lic a
cids
, and
thei
r anh
ydrid
es, h
alid
es, e
tc.
21 O
ther
inor
gani
c ch
emic
als
22 C
arbo
xylic
aci
ds, a
nd th
eir a
nhyd
rides
, hal
ides
, etc
. 23
Mis
cella
neou
s arti
cles
of m
ater
ials
of d
ivis
ion
58
24 O
rgan
o-in
orga
nic
and
hete
rocy
clic
com
poun
ds
25 C
arbo
xylic
aci
ds, a
nd th
eir a
nhyd
rides
, hal
ides
, etc
. 26
Car
boxy
lic a
cids
, and
thei
r anh
ydrid
es, h
alid
es, e
tc.
27 U
nive
rsal
s, pl
ates
and
shee
ts o
f iro
n or
stee
l; in
gots
and
oth
er p
rimar
y fo
rms o
f iro
n or
stee
l 28
Ang
les,
shap
es &
sect
ions
& sh
eet p
iling
of i
ron
or st
eel;
stru
ctur
es &
par
ts o
f stru
ctur
es, i
ron
or st
eel;
plat
es
29 S
eam
less
tube
s and
pip
es; b
lank
s for
tube
s & p
ipes
30
Sea
mle
ss tu
bes a
nd p
ipes
; bla
nks f
or tu
bes &
pip
es
31 R
efin
ed su
gars
and
oth
er p
rodu
cts o
f ref
ined
bee
t or c
ane
32 V
eget
able
text
ile fi
bers
and
was
te o
f suc
h fib
ers
33 P
aper
and
pap
erbo
ard,
cut
to si
ze o
r sha
pe,
n.e.
s.; p
aper
and
pap
erbo
ard,
in ro
lls o
r she
ets,
n.e.
s. 34
Pro
vita
min
s and
vita
min
s, na
tura
l or p
rodu
ced
by sy
nthe
sis
35 P
umps
and
com
pres
sors
, fan
s & b
low
ers,
cent
rifug
es
36 O
ther
inor
gani
c ch
emic
als
63
TA
BL
E 1
1
IMPO
RT
AN
CE
OF
CA
RT
EL
IZE
D G
OO
DS
IN D
EV
EL
OPI
NG
CO
UN
TR
Y IM
POR
TS
(4-D
IGIT
SIT
C),
1997
1
Low
Inco
me
Coun
tries
Lo
wer M
iddl
e In
com
e Co
untri
es
Upp
er M
iddl
e In
com
e Co
untri
es
All D
evel
opin
g Co
untri
es
Hig
h In
com
e Co
untri
es
Prod
uct
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Alu
min
um P
hosp
hide
2
0.05
1%
0.01
8%
0.03
2%
0.02
9%
0.03
5%
Bee
r3 0.
060%
0.
056%
0.
031%
0.
043%
0.
096%
B
rom
ine4
0.10
9%
0.10
8%
0.15
1%
0.13
1%
0.15
9%
Cab
le-S
taye
d B
ridge
s5 N
A
NA
N
A
NA
N
A
Car
bon
Cat
hode
Blo
ck6
0.45
1%
0.84
8%
1.11
8%
0.95
0%
0.72
7%
Car
bonl
ess P
aper
7 0.
104%
0.
094%
0.
102%
0.
099%
0.
111%
C
arto
nboa
rd8
0.13
1%
0.13
1%
0.21
8%
0.17
8%
0.13
3%
Cem
ent9
0.36
7%
0.09
8%
0.10
8%
0.13
3%
0.07
1%
Citr
ic A
cid10
N
A
NA
N
A
NA
N
A
Expl
osiv
es11
N
A
NA
N
A
NA
N
A
Ferr
osili
con12
0.
080%
0.
028%
0.
098%
0.
071%
0.
122%
Fi
ne A
rts13
0.
019%
0.
019%
0.
034%
0.
027%
0.
181%
G
raph
ite E
lect
rode
s14
0.45
1%
0.84
8%
1.11
8%
0.95
0%
0.72
7%
Isos
tatic
Gra
phite
15
0.97
5%
0.86
0%
0.97
9%
0.93
6%
0.74
2%
Lam
inat
ed P
last
ic T
ubes
16
0.35
8%
0.55
5%
0.92
0%
0.72
9%
0.73
0%
Lysi
ne17
N
A
NA
N
A
NA
N
A
Mal
tol18
N
A
NA
N
A
NA
N
A
Mar
ine
Con
stru
ctio
n19
NA
N
A
NA
N
A
NA
64
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Perc
ent o
f tot
al
impo
rts
Mon
ochl
oroa
cetic
Aci
d20
NA
N
A
NA
N
A
NA
N
ucle
otid
es21
N
A
NA
N
A
NA
N
A
Org
anic
Per
oxid
es22
N
A
NA
N
A
NA
N
A
Plas
tic D
inne
rwar
e23
0.35
8%
0.55
5%
0.92
0%
0.72
9%
0.73
0%
Sodi
um E
ryth
orba
te24
N
A
NA
N
A
NA
N
A
Sodi
um G
luco
nate
25
NA
N
A
NA
N
A
NA
So
rbat
es26
N
A
NA
N
A
NA
N
A
Stai
nles
s Ste
el27
N
A
NA
N
A
NA
N
A
Stee
l Bea
m28
0.
172%
0.
158%
0.
149%
0.
155%
0.
119%
St
eel H
eatin
g Pi
pe29
0.
429%
0.
299%
0.
255%
0.
289%
0.
121%
St
eel T
ube,
Sea
mle
ss30
0.
429%
0.
299%
0.
255%
0.
289%
0.
121%
Su
gar31
0.
831%
0.
272%
0.
060%
0.
219%
0.
058%
Ta
mpi
co F
iber
32
NA
N
A
NA
N
A
NA
Th
erm
al F
ax P
aper
33
0.40
1%
0.49
1%
0.33
1%
0.39
6%
0.41
5%
Vita
min
s34
0.06
6%
0.08
1%
0.06
5%
0.07
1%
0.06
9%
Was
tew
ater
Con
stru
ctio
n35
NA
N
A
NA
N
A
NA
Zi
nc P
hosp
hate
36
NA
N
A
NA
N
A
NA
To
tal
4.96
1%
4.67
0%
5.57
1%
5.18
4%
4.61
8%
65
TA
BL
E 1
2 C
AR
TE
LIZ
ED
GO
OD
S A
S A
SH
AR
E O
F D
EV
EL
OPI
NG
CO
UN
TR
Y G
DP
(4-D
IGIT
SIT
C),
1997
1
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Perc
ent o
f GD
P Pe
rcen
t of G
DP
Perc
ent o
f GD
P Pe
rcen
t of G
DP
Perc
ent o
f GD
P
Alu
min
um P
hosp
hide
2 0.
0078
%
0.00
48%
0.
0074
%
0.00
67%
0.
0065
%
Bee
r3 0.
0091
%
0.01
47%
0.
0072
%
0.00
98%
0.
0182
%
Bro
min
e4 0.
0166
%
0.02
84%
0.
0346
%
0.02
98%
0.
0300
%
Cab
le-S
taye
d B
ridge
s5 N
A
NA
N
A
NA
N
A
Car
bon
Cat
hode
Blo
ck6
0.06
86%
0.
2232
%
0.25
72%
0.
2163
%
0.13
72%
C
arbo
nles
s Pap
er7
0.01
58%
0.
0248
%
0.02
34%
0.
0226
%
0.02
10%
C
arto
nboa
rd8
0.02
00%
0.
0345
%
0.05
02%
0.
0405
%
0.02
50%
C
emen
t9 0.
0559
%
0.02
58%
0.
0249
%
0.03
02%
0.
0134
%
Citr
ic A
cid10
N
A
NA
N
A
NA
N
A
Expl
osiv
es11
N
A
NA
N
A
NA
N
A
Ferr
osili
con12
0.
0122
%
0.00
73%
0.
0226
%
0.01
62%
0.
0230
%
Fine
Arts
13
0.00
28%
0.
0050
%
0.00
78%
0.
0061
%
0.03
42%
G
raph
ite E
lect
rode
s14
0.06
86%
0.
2232
%
0.25
72%
0.
2163
%
0.13
72%
Is
osta
tic G
raph
ite15
0.
1484
%
0.22
65%
0.
2252
%
0.21
32%
0.
1401
%
Lam
inat
ed P
last
ic T
ubes
16
0.05
44%
0.
1461
%
0.21
15%
0.
1660
%
0.13
77%
Ly
sine
17
NA
N
A
NA
N
A
NA
M
alto
l18
NA
N
A
NA
N
A
NA
M
arin
e C
onst
ruct
ion19
N
A
NA
N
A
NA
N
A
66
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Perc
ent o
f GD
P Pe
rcen
t of G
DP
Perc
ent o
f GD
P Pe
rcen
t of G
DP
Perc
ent o
f GD
P
Mon
ochl
oroa
cetic
Aci
d20
NA
N
A
NA
N
A
NA
N
ucle
otid
es21
N
A
NA
N
A
NA
N
A
Org
anic
Per
oxid
es22
N
A
NA
N
A
NA
N
A
Plas
tic D
inne
rwar
e23
0.05
44%
0.
1461
%
0.21
15%
0.
1660
%
0.13
77%
So
dium
Ery
thor
bate
24
NA
N
A
NA
N
A
NA
So
dium
Glu
cona
te25
N
A
NA
N
A
NA
N
A
Sorb
ates
26
NA
N
A
NA
N
A
NA
St
ainl
ess S
teel
27
NA
N
A
NA
N
A
NA
St
eel B
eam
28
0.02
62%
0.
0416
%
0.03
43%
0.
0353
%
0.02
25%
St
eel H
eatin
g Pi
pe29
0.
0654
%
0.07
87%
0.
0586
%
0.06
59%
0.
0228
%
Stee
l Tub
e, S
eam
less
30
0.06
54%
0.
0787
%
0.05
86%
0.
0659
%
0.02
28%
Su
gar31
0.
1265
%
0.07
18%
0.
0138
%
0.04
98%
0.
0109
%
Tam
pico
Fib
er32
N
A
NA
N
A
NA
N
A
Ther
mal
Fax
Pap
er33
0.
0611
%
0.12
94%
0.
0761
%
0.09
01%
0.
0783
%
Vita
min
s34
0.01
00%
0.
0212
%
0.01
49%
0.
0161
%
0.01
30%
W
aste
wat
er C
onst
ruct
ion35
N
A
NA
N
A
NA
N
A
Zinc
Pho
spha
te36
N
A
NA
N
A
NA
N
A
Tota
l 0.
7552
%
1.22
98%
1.
2812
%
1.18
05%
0.
8715
%
67
TA
BL
E 1
3
TO
TA
L V
AL
UE
OF
CA
RT
EL
-AFF
EC
TE
D IM
POR
TS
(4-D
IGIT
SIT
C),
1997
1
Low
Inco
me
Coun
tries
Lo
wer M
iddl
e In
com
e Co
untri
es
Upp
er M
iddl
e In
com
e Co
untri
es
All D
evel
opin
g Co
untri
es
Hig
h In
com
e Co
untri
es
Prod
uct
Val
ue o
f tot
al
impo
rts
($00
0)
Val
ue o
f tot
al
impo
rts
($00
0)
Val
ue o
f tot
al
impo
rts
($00
0)
Tot
al Im
port
s by
Dev
elop
ing
Cou
ntri
es
($00
0)
Val
ue o
f tot
al
impo
rts
($00
0)
Alu
min
um P
hosp
hide
2 70
,402
82
,664
22
0,86
0 37
3,92
6 1,
452,
864
Bee
r3 82
,321
25
4,61
2 21
4,36
1 55
1,29
4 4,
053,
841
Bro
min
e4 15
0,14
8 49
1,90
7 1,
031,
786
1,67
3,84
1 6,
673,
615
Cab
le-S
taye
d B
ridge
s5 N
A
NA
N
A
NA
N
A
Car
bon
Cat
hode
Blo
ck6
621,
595
3,85
9,81
5 7,
659,
252
12,1
40,6
62
30,5
57,5
46
Car
bonl
ess P
aper
7 14
2,81
3 42
8,32
7 69
6,82
1 1,
267,
961
4,68
2,64
5 C
arto
nboa
rd8
181,
251
595,
795
1,49
5,83
8 2,
272,
884
5,57
6,24
7 C
emen
t9 50
6,33
9 44
5,98
4 74
1,97
9 1,
694,
302
2,98
4,99
1 C
itric
Aci
d10
NA
N
A
NA
N
A
NA
Ex
plos
ives
11
NA
N
A
NA
N
A
NA
Fe
rros
ilico
n12
110,
896
126,
282
672,
946
910,
124
5,13
0,07
8 Fi
ne A
rts13
25
,785
86
,150
23
1,26
7 34
3,20
2 7,
619,
113
Gra
phite
Ele
ctro
des14
62
1,59
5 3,
859,
815
7,65
9,25
2 12
,140
,662
30
,557
,546
Is
osta
tic G
raph
ite15
1,
344,
868
3,91
6,39
2 6,
707,
800
11,9
69,0
60
31,2
15,2
17
Lam
inat
ed P
last
ic T
ubes
16
493,
255
2,52
5,85
6 6,
297,
960
9,31
7,07
1 30
,679
,451
68
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Val
ue o
f tot
al
impo
rts
($00
0)
Val
ue o
f tot
al
impo
rts
($00
0)
Val
ue o
f tot
al
impo
rts
($00
0)
Tot
al Im
port
s by
Dev
elop
ing
Cou
ntri
es
($00
0)
Val
ue o
f tot
al
impo
rts
($00
0)
Lysi
ne17
N
A
NA
N
A
NA
N
A
Mal
tol18
N
A
NA
N
A
NA
N
A
Mar
ine
Con
stru
ctio
n19
NA
N
A
NA
N
A
NA
M
onoc
hlor
oace
tic A
cid20
N
A
NA
N
A
NA
N
A
Nuc
leot
ides
21
NA
N
A
NA
N
A
NA
O
rgan
ic P
erox
ides
22
NA
N
A
NA
N
A
NA
Pl
astic
Din
nerw
are23
49
3,25
5 2,
525,
856
6,29
7,96
0 9,
317,
071
30,6
79,4
51
Sodi
um E
ryth
orba
te24
N
A
NA
N
A
NA
N
A
Sodi
um G
luco
nate
25
NA
N
A
NA
N
A
NA
So
rbat
es26
N
A
NA
N
A
NA
N
A
Stai
nles
s Ste
el27
N
A
NA
N
A
NA
N
A
Stee
l Bea
m28
23
6,97
6 71
9,38
3 1,
022,
477
1,97
8,83
6 5,
016,
869
Stee
l Hea
ting
Pipe
29
592,
196
1,36
1,42
2 1,
746,
713
3,70
0,33
1 5,
090,
477
Stee
l Tub
e, S
eam
less
30
592,
196
1,36
1,42
2 1,
746,
713
3,70
0,33
1 5,
090,
477
Suga
r31
1,14
6,44
7 1,
240,
914
410,
048
2,79
7,40
9 2,
424,
484
Tam
pico
Fib
er32
N
A
NA
N
A
NA
N
A
Ther
mal
Fax
Pap
er33
55
3,33
5 2,
237,
895
2,26
5,39
4 5,
056,
624
17,4
37,6
72
Vita
min
s34
90,9
85
367,
034
444,
804
902,
823
2,88
7,21
6 W
aste
wat
er C
onst
ruct
ion35
N
A
NA
N
A
NA
N
A
69
Lo
w In
com
e Co
untri
es
Lowe
r Mid
dle
Inco
me
Coun
tries
U
pper
Mid
dle
Inco
me
Coun
tries
Al
l Dev
elop
ing
Coun
tries
H
igh
Inco
me
Coun
tries
Prod
uct
Val
ue o
f tot
al
impo
rts
($00
0)
Val
ue o
f tot
al
impo
rts
($00
0)
Val
ue o
f tot
al
impo
rts
($00
0)
Tot
al Im
port
s by
Dev
elop
ing
Cou
ntri
es
($00
0)
Val
ue o
f tot
al
impo
rts
($00
0)
Zinc
Pho
spha
te36
N
A
NA
N
A
NA
N
A
Tota
l 6,
168,
361
18,1
44,6
37
30,3
64,4
68
54,6
77,4
66
157,
906,
079
Not
es to
Tab
les 1
1, 1
2, a
nd 1
3:
1 The
list
of d
evel
opin
g co
untri
es is
take
n fr
om W
orld
Dev
elop
men
t Rep
ort 2
000/
2001
: At
tack
ing
Pove
rty
(Wor
ld B
ank)
, pp.
334
-35.
A
s sta
ted
on p
. 335
“Lo
w in
com
e an
d m
iddl
e-in
com
e ec
onom
ies a
re so
met
imes
refe
rred
to a
s dev
elop
ing
coun
tries
.”
2 Inor
gani
c ch
emic
al p
rodu
cts,
n.e.
s 3 B
eer m
ade
from
mal
t (in
clud
ing
ale,
stou
t and
por
ter)
4 P
heno
ls a
nd p
heny
l alc
ohol
s and
thei
r hal
ogen
ated
der
ivat
ives
; eth
ers,
alco
hol p
erox
ides
, eth
er p
erox
ides
, epo
xide
s etc
; hal
ogen
ated
de
rivat
ives
of h
ydro
carb
ons
5 Stru
ctur
es &
par
ts o
f stru
ctur
es: i
ron/
stee
l pla
tes;
stru
ctur
es a
nd p
arts
of s
truct
ures
: alu
min
um p
late
s and
rods
6 O
ther
ele
ctric
al m
achi
nery
and
equ
ipm
ent
7 Pap
er a
nd p
aper
boar
d, c
ut to
size
or s
hape
, n.e
.s.
8 Box
es, b
ags &
oth
er p
acki
ng c
onta
iner
s, of
pap
er o
r pap
erbo
ard
9 Por
tland
cem
ent,
cem
ent f
ondu
, sla
g ce
men
t, et
c.
10 C
arbo
xylic
aci
ds w
ith a
lcoh
ol, p
heno
l, et
c. fu
nctio
n 11
Exp
losi
ves a
nd p
yrot
echn
ic p
rodu
cts
12 F
erro
-allo
ys
13 A
rt, c
olle
ctor
s’ p
iece
s and
ant
ique
s 14
Oth
er e
lect
rical
mac
hine
ry a
nd e
quip
men
t 15
Org
anic
che
mic
al p
rodu
cts,
n.e.
s.; c
hem
ical
pro
duct
s and
pre
para
tions
, n.e
.s.
16 M
isce
llane
ous a
rticl
es o
f mat
eria
ls o
f div
isio
n.58
; acr
ylic
pol
ymer
s, m
etha
cryl
ic p
olym
ers,
etc.
; am
inop
last
s 17
Sing
le o
r com
plex
oxy
gen-
func
tion
amin
o-co
mpo
unds
18
Het
eroc
yclic
com
poun
ds; n
ucle
ic a
cids
19
Tug
s, sp
ecia
l pur
pose
ves
sels
, flo
atin
g st
ruct
ures
20
Mon
ocar
boxy
lic a
cids
and
thei
r anh
ydrid
es, h
alid
es
70
21 M
etal
lic sa
lts a
nd p
erox
ysal
ts o
f ino
rgan
ic a
cids
22
Mon
ocar
boxy
lic a
cids
& th
eir a
nhyd
rides
, hal
ides
; pol
ycar
boxy
lic a
cids
and
thei
r anh
ydrid
es, e
tc.;
carb
oxyl
ic a
cids
with
alc
ohol
, ph
enol
, etc
. fun
ctio
n 23
Mis
cella
neou
s arti
cles
of m
ater
ials
of d
ivis
ion
58
24 H
eter
ocyc
lic c
ompo
unds
; nuc
leic
aci
ds
25 C
arbo
xylic
aci
ds w
ith a
lcoh
ol, p
heno
l, et
c. fu
nctio
n 26
Mon
ocar
boxy
lic a
cids
and
thei
r anh
ydrid
es, h
alid
es
27 S
heet
s and
pla
tes,
rolle
d >
4.75
mm
of i
ron/
stee
l; sh
eets
and
pla
tes,
rolle
d; th
ickn
ess o
f les
s tha
n 3m
m; s
heet
s and
pla
tes,
rolle
d th
ickn
ess 3
mm
to 4
.75
mm
iron
or s
teel
; iro
n or
stee
l coi
ls fo
r re-
rolli
ng
28 A
ngle
s, sh
apes
& se
ctio
ns &
shee
t pili
ng o
f iro
n or
stee
l; st
ruct
ures
& p
arts
of s
truct
ures
, iro
n or
stee
l; pl
ates
29
Sea
mle
ss tu
bes a
nd p
ipes
; bla
nks f
or tu
bes &
pip
es
30 S
eam
less
tube
s and
pip
es; b
lank
s for
tube
s & p
ipes
31
Ref
ined
suga
rs a
nd o
ther
pro
duct
s of r
efin
ed b
eet o
r can
e 32
Sisa
l and
oth
er fi
bers
of a
gave
fam
ily, r
aw o
r pro
cess
ed
33 P
aper
and
pap
erbo
ard,
cut
to si
ze o
r sha
pe,
n.e.
s.; p
aper
and
pap
erbo
ard,
in ro
lls o
r she
ets,
n.e.
s. 34
Pro
vita
min
s and
vita
min
s, na
tura
l or p
rodu
ced
by sy
nthe
sis
35 F
ilter
ing
and
purif
ying
mac
hine
ry fo
r liq
uids
and
gas
es
36 M
etal
lic sa
lts a
nd p
erox
ysal
ts o
f ino
rgan
ic a
cids
71
FIG
UR
E 1
Cit
ric
Aci
d P
rice
s, 1
990
- 19
99(C
arte
l: 1
991
- 19
95)
Com
paris
on: C
MR
v P
M S
pot A
vg P
rice
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
Apr-90
Jul-90
Oct-90
Jan-91
Apr-91
Jul-91
Oct-91
Jan-92
Apr-92
Jul-92
Oct-92
Jan-93
Apr-93
Jul-93
Oct-93
Jan-94
Apr-94
Jul-94
Oct-94
Jan-95
Apr-95
Jul-95
Oct-95
Jan-96
Apr-96
Jul-96
Oct-96
Jan-97
Apr-97
Jul-97
Oct-97
Jan-98
Apr-98
Jul-98
Oct-98
Jan-99
Apr-99
Jul-99
Oct-99
Yea
r
US
D c
ents
per
lb
CM
R P
rice
FM
Spo
t Avg
Pric
eSources:
Che
mic
al M
arke
t Rep
orte
r (C
MR
), P
urch
asin
g M
agaz
ine
(PM
)
US
P, a
nhyd
, Gra
n. 1
00 lb
bgs
PM S
pot A
vg P
rice
72
FIG
UR
E 2
Gra
phite
Ele
ctro
de P
rice
s, 1
980
- 200
0(C
arte
l: Ju
ly 1
992
- Jun
e 19
97)
$0
$500
$1,0
00
$1,5
00
$2,0
00
$2,5
00
$3,0
00
$3,5
00Q1'80
Q1'81
Q1'89
Q1'90
Q2'92
Q1'93
Q2'93
Q1'94
Q3'94
Q1'95
Q2'95
Q4'95
Q1'96
Q2'96
Q3'96
Q1'97
Q2'97
Q3'97
Q4'97
Q1'98
Q3'98
Q4'98
Q2'99
Q3'99
Q4'99
Q1'00
Q2'00
Q3'00
Yea
r
US
D p
er M
etri
c T
on
Sour
ces:
DO
J Sen
tenc
ing
Mem
oran
dum
of O
ctob
er 1
9, 1
999,
and
a v
arie
ty o
f oth
er so
urce
s, in
clud
ing
Forb
es, O
il an
d G
as J
ourn
al, D
ow J
ones
C
omm
odity
Ser
vice
, UC
AR
ear
ning
s rep
orts
, and
C/G
SEC
filin
gs.
73
FIG
UR
E 3
O
CT
G P
rice
s (C
arte
l: 1
990-
95)
OC
TG S
POT
MAR
KET
PR
ICES
500
550
600
650
700
750
800
850
900
950
1000
1050
1100
'86
'87
'88
'89
'90
'91
'92
'93
'94
'95
$ / T
onSE
AMLE
SS &
ER
W IT
EMS
AVER
AGE
PRIC
ES
Sour
ce: P
ipe
Logi
x, In
c.
All
ER
W It
ems
>>
<<<
All
Seam
less
Item
s '96
'97
'98
74
FIG
UR
E 4
Car
tel S
hare
of S
teel
Tub
e E
xpor
ts
010203040506070
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
Yea
rs
Percentage Share of Exports
Car
tel S
hare
of S
teel
Tub
e E
xpor
ts
So
urce
: R
ober
t Fee
nstra
’s W
orld
Tra
de F
low
s, 19
80-1
997,
With
Pro
duct
ion
And
Tari
ff D
ata
(199
9).